identifier
stringlengths 7
18
| space
stringclasses 4
values | uid
stringlengths 1
6
| arch_str
stringlengths 1
32
| input
stringlengths 8.51k
461k
| target_metric
stringclasses 1
value | val_accuracy
float64 0
95.1
| flops
float64 31.1M
14.7B
| params
float64 227k
50M
| metadata
stringlengths 0
1.46k
| metainformation
stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|
FBNet_1391
|
FBNet
|
1391
|
1391
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_688[FLOAT, 16x3x3x3]
%onnx::Conv_689[FLOAT, 16]
%onnx::Conv_691[FLOAT, 16x8x1x1]
%onnx::Conv_694[FLOAT, 16x1x3x3]
%onnx::Conv_697[FLOAT, 16x8x1x1]
%onnx::Conv_700[FLOAT, 48x16x1x1]
%onnx::Conv_701[FLOAT, 48]
%onnx::Conv_703[FLOAT, 48x1x5x5]
%onnx::Conv_706[FLOAT, 24x48x1x1]
%onnx::Conv_707[FLOAT, 24]
%onnx::Conv_709[FLOAT, 72x24x1x1]
%onnx::Conv_710[FLOAT, 72]
%onnx::Conv_712[FLOAT, 72x1x5x5]
%onnx::Conv_715[FLOAT, 24x72x1x1]
%onnx::Conv_718[FLOAT, 144x24x1x1]
%onnx::Conv_719[FLOAT, 144]
%onnx::Conv_721[FLOAT, 144x1x3x3]
%onnx::Conv_724[FLOAT, 24x144x1x1]
%onnx::Conv_727[FLOAT, 72x24x1x1]
%onnx::Conv_730[FLOAT, 72x1x5x5]
%onnx::Conv_733[FLOAT, 24x72x1x1]
%onnx::Conv_736[FLOAT, 24x12x1x1]
%onnx::Conv_739[FLOAT, 24x1x3x3]
%onnx::Conv_742[FLOAT, 32x12x1x1]
%onnx::Conv_743[FLOAT, 32]
%onnx::Conv_745[FLOAT, 192x32x1x1]
%onnx::Conv_746[FLOAT, 192]
%onnx::Conv_748[FLOAT, 192x1x5x5]
%onnx::Conv_751[FLOAT, 32x192x1x1]
%onnx::Conv_754[FLOAT, 192x32x1x1]
%onnx::Conv_757[FLOAT, 192x1x3x3]
%onnx::Conv_760[FLOAT, 32x192x1x1]
%onnx::Conv_763[FLOAT, 32x16x1x1]
%onnx::Conv_766[FLOAT, 32x1x3x3]
%onnx::Conv_769[FLOAT, 32x16x1x1]
%onnx::Conv_772[FLOAT, 64x32x1x1]
%onnx::Conv_773[FLOAT, 64]
%onnx::Conv_775[FLOAT, 384x64x1x1]
%onnx::Conv_776[FLOAT, 384]
%onnx::Conv_778[FLOAT, 384x1x5x5]
%onnx::Conv_781[FLOAT, 64x384x1x1]
%onnx::Conv_784[FLOAT, 192x64x1x1]
%onnx::Conv_787[FLOAT, 192x1x5x5]
%onnx::Conv_790[FLOAT, 64x192x1x1]
%onnx::Conv_793[FLOAT, 64x32x1x1]
%onnx::Conv_796[FLOAT, 64x1x5x5]
%onnx::Conv_799[FLOAT, 64x32x1x1]
%onnx::Conv_802[FLOAT, 64x32x1x1]
%onnx::Conv_805[FLOAT, 64x1x5x5]
%onnx::Conv_808[FLOAT, 112x32x1x1]
%onnx::Conv_809[FLOAT, 112]
%onnx::Conv_811[FLOAT, 336x112x1x1]
%onnx::Conv_812[FLOAT, 336]
%onnx::Conv_814[FLOAT, 336x1x5x5]
%onnx::Conv_817[FLOAT, 112x336x1x1]
%onnx::Conv_820[FLOAT, 112x112x1x1]
%onnx::Conv_823[FLOAT, 112x1x5x5]
%onnx::Conv_826[FLOAT, 112x112x1x1]
%onnx::Conv_829[FLOAT, 112x112x1x1]
%onnx::Conv_832[FLOAT, 112x1x5x5]
%onnx::Conv_835[FLOAT, 112x112x1x1]
%onnx::Conv_838[FLOAT, 112x112x1x1]
%onnx::Conv_841[FLOAT, 112x1x3x3]
%onnx::Conv_844[FLOAT, 184x112x1x1]
%onnx::Conv_845[FLOAT, 184]
%onnx::Conv_847[FLOAT, 552x184x1x1]
%onnx::Conv_848[FLOAT, 552]
%onnx::Conv_850[FLOAT, 552x1x5x5]
%onnx::Conv_853[FLOAT, 184x552x1x1]
%onnx::Conv_856[FLOAT, 184x184x1x1]
%onnx::Conv_859[FLOAT, 184x1x5x5]
%onnx::Conv_862[FLOAT, 184x184x1x1]
%onnx::Conv_865[FLOAT, 184x92x1x1]
%onnx::Conv_868[FLOAT, 184x1x3x3]
%onnx::Conv_871[FLOAT, 352x92x1x1]
%onnx::Conv_872[FLOAT, 352]
%onnx::Conv_874[FLOAT, 1504x352x1x1]
%onnx::Conv_875[FLOAT, 1504]
) {
%onnx::Conv_869 = Identity(%onnx::Conv_845)
%onnx::Conv_866 = Identity(%onnx::Conv_845)
%onnx::Conv_863 = Identity(%onnx::Conv_845)
%onnx::Conv_860 = Identity(%onnx::Conv_845)
%onnx::Conv_857 = Identity(%onnx::Conv_845)
%onnx::Conv_854 = Identity(%onnx::Conv_845)
%onnx::Conv_851 = Identity(%onnx::Conv_848)
%onnx::Conv_842 = Identity(%onnx::Conv_809)
%onnx::Conv_839 = Identity(%onnx::Conv_809)
%onnx::Conv_836 = Identity(%onnx::Conv_809)
%onnx::Conv_833 = Identity(%onnx::Conv_809)
%onnx::Conv_830 = Identity(%onnx::Conv_809)
%onnx::Conv_827 = Identity(%onnx::Conv_809)
%onnx::Conv_824 = Identity(%onnx::Conv_809)
%onnx::Conv_821 = Identity(%onnx::Conv_809)
%onnx::Conv_818 = Identity(%onnx::Conv_809)
%onnx::Conv_815 = Identity(%onnx::Conv_812)
%onnx::Conv_806 = Identity(%onnx::Conv_773)
%onnx::Conv_803 = Identity(%onnx::Conv_773)
%onnx::Conv_800 = Identity(%onnx::Conv_773)
%onnx::Conv_797 = Identity(%onnx::Conv_773)
%onnx::Conv_794 = Identity(%onnx::Conv_773)
%onnx::Conv_791 = Identity(%onnx::Conv_773)
%onnx::Conv_788 = Identity(%onnx::Conv_746)
%onnx::Conv_785 = Identity(%onnx::Conv_746)
%onnx::Conv_782 = Identity(%onnx::Conv_773)
%onnx::Conv_779 = Identity(%onnx::Conv_776)
%onnx::Conv_770 = Identity(%onnx::Conv_743)
%onnx::Conv_767 = Identity(%onnx::Conv_743)
%onnx::Conv_764 = Identity(%onnx::Conv_743)
%onnx::Conv_761 = Identity(%onnx::Conv_743)
%onnx::Conv_758 = Identity(%onnx::Conv_746)
%onnx::Conv_755 = Identity(%onnx::Conv_746)
%onnx::Conv_752 = Identity(%onnx::Conv_743)
%onnx::Conv_749 = Identity(%onnx::Conv_746)
%onnx::Conv_740 = Identity(%onnx::Conv_707)
%onnx::Conv_737 = Identity(%onnx::Conv_707)
%onnx::Conv_734 = Identity(%onnx::Conv_707)
%onnx::Conv_731 = Identity(%onnx::Conv_710)
%onnx::Conv_728 = Identity(%onnx::Conv_710)
%onnx::Conv_725 = Identity(%onnx::Conv_707)
%onnx::Conv_722 = Identity(%onnx::Conv_719)
%onnx::Conv_716 = Identity(%onnx::Conv_707)
%onnx::Conv_713 = Identity(%onnx::Conv_710)
%onnx::Conv_704 = Identity(%onnx::Conv_701)
%onnx::Conv_698 = Identity(%onnx::Conv_689)
%onnx::Conv_695 = Identity(%onnx::Conv_689)
%onnx::Conv_692 = Identity(%onnx::Conv_689)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_688, %onnx::Conv_689)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.8/Add_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.9/relu/Relu_output_0 = Relu(%/cells.9/conv/Conv_output_0)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/relu/Relu_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/relu/Relu_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_823, %onnx::Conv_824)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_826, %onnx::Conv_827)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_829, %onnx::Conv_830)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_832, %onnx::Conv_833)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_835, %onnx::Conv_836)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_838, %onnx::Conv_839)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_841, %onnx::Conv_842)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_844, %onnx::Conv_845)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_847, %onnx::Conv_848)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_850, %onnx::Conv_851)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_853, %onnx::Conv_854)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_856, %onnx::Conv_857)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_859, %onnx::Conv_860)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_862, %onnx::Conv_863)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_865, %onnx::Conv_866)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_868, %onnx::Conv_869)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_871, %onnx::Conv_872)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_874, %onnx::Conv_875)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%686 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %686
}
|
val_accuracy
| 0
| 66,162,560
| 1,371,892
|
{'zcp_synflow': 77.25158005374898, 'zcp_zen': 66.75434875488281, 'zcp_epe_nas': 16.12152486233673, 'zcp_fisher': 0.10830622166395187, 'zcp_flops': 66162560.0, 'zcp_grad_norm': 24.301162719726562, 'zcp_grasp': -0.17497825622558594, 'zcp_jacov': -16.057567901877782, 'zcp_l2_norm': 576.9022216796875, 'zcp_nwot': 214.60446623412577, 'zcp_params': 1371892.0, 'zcp_plain': 0.005425637122243643, 'zcp_snip': 42.44517135620117, 'lat_1080ti_1': 0.6525984134156153, 'lat_1080ti_32': 0.6664759621979338, 'lat_1080ti_64': 0.6095184771173432, 'lat_2080ti_1': 0.6764042969231903, 'lat_2080ti_32': 0.6397503469757887, 'lat_2080ti_64': 0.6238385834001589, 'lat_essential_ph_1': 0.3584905660377358, 'lat_eyeriss': 0.45912566702730445, 'lat_fpga': 0.4168258146231409, 'lat_gold_6226': 0.2260811731761784, 'lat_gold_6240': 0.3582752832792652, 'lat_pixel2': 0.45652173913043476, 'lat_pixel3': 0.4875519153020113, 'lat_raspi4': 0.42596179115862176, 'lat_samsung_a50': 0.16842105263157894, 'lat_samsung_s7': 0.13385826771653545, 'lat_silver_4114': 0.3963984247206506, 'lat_silver_4210r': 0.4284804848596081, 'lat_titan_rtx_1': 0.6468125708177124, 'lat_titan_rtx_32': 0.6384056337167493, 'lat_titan_rtx_64': 0.6501521316836746, 'lat_titanx_1': 0.33466795797628374, 'lat_titanx_32': 0.635249380178132, 'lat_titanx_64': 0.6500330591029904, 'lat_titanxp_1': 0.6341651709221119, 'lat_titanxp_32': 0.648455555003125, 'lat_titanxp_64': 0.6179231715747348}
| |
FBNet_2404
|
FBNet
|
2404
|
2404
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_704[FLOAT, 16x3x3x3]
%onnx::Conv_705[FLOAT, 16]
%onnx::Conv_707[FLOAT, 48x16x1x1]
%onnx::Conv_708[FLOAT, 48]
%onnx::Conv_710[FLOAT, 48x1x5x5]
%onnx::Conv_713[FLOAT, 16x48x1x1]
%onnx::Conv_716[FLOAT, 16x16x1x1]
%onnx::Conv_719[FLOAT, 16x1x3x3]
%onnx::Conv_722[FLOAT, 24x16x1x1]
%onnx::Conv_723[FLOAT, 24]
%onnx::Conv_725[FLOAT, 144x24x1x1]
%onnx::Conv_726[FLOAT, 144]
%onnx::Conv_728[FLOAT, 144x1x5x5]
%onnx::Conv_731[FLOAT, 24x144x1x1]
%onnx::Conv_734[FLOAT, 24x24x1x1]
%onnx::Conv_737[FLOAT, 24x1x3x3]
%onnx::Conv_740[FLOAT, 24x24x1x1]
%onnx::Conv_743[FLOAT, 24x24x1x1]
%onnx::Conv_746[FLOAT, 24x1x3x3]
%onnx::Conv_749[FLOAT, 24x24x1x1]
%onnx::Conv_752[FLOAT, 72x24x1x1]
%onnx::Conv_753[FLOAT, 72]
%onnx::Conv_755[FLOAT, 72x1x3x3]
%onnx::Conv_758[FLOAT, 32x72x1x1]
%onnx::Conv_759[FLOAT, 32]
%onnx::Conv_761[FLOAT, 32x16x1x1]
%onnx::Conv_764[FLOAT, 32x1x5x5]
%onnx::Conv_767[FLOAT, 32x16x1x1]
%onnx::Conv_770[FLOAT, 32x16x1x1]
%onnx::Conv_773[FLOAT, 32x1x3x3]
%onnx::Conv_776[FLOAT, 32x16x1x1]
%onnx::Conv_779[FLOAT, 192x32x1x1]
%onnx::Conv_780[FLOAT, 192]
%onnx::Conv_782[FLOAT, 192x1x3x3]
%onnx::Conv_785[FLOAT, 32x192x1x1]
%onnx::Conv_788[FLOAT, 96x32x1x1]
%onnx::Conv_789[FLOAT, 96]
%onnx::Conv_791[FLOAT, 96x1x5x5]
%onnx::Conv_794[FLOAT, 64x96x1x1]
%onnx::Conv_795[FLOAT, 64]
%onnx::Conv_797[FLOAT, 64x32x1x1]
%onnx::Conv_800[FLOAT, 64x1x3x3]
%onnx::Conv_803[FLOAT, 64x32x1x1]
%onnx::Conv_806[FLOAT, 64x32x1x1]
%onnx::Conv_809[FLOAT, 64x1x3x3]
%onnx::Conv_812[FLOAT, 64x32x1x1]
%onnx::Conv_815[FLOAT, 64x64x1x1]
%onnx::Conv_818[FLOAT, 64x1x3x3]
%onnx::Conv_821[FLOAT, 112x64x1x1]
%onnx::Conv_822[FLOAT, 112]
%onnx::Conv_824[FLOAT, 336x112x1x1]
%onnx::Conv_825[FLOAT, 336]
%onnx::Conv_827[FLOAT, 336x1x5x5]
%onnx::Conv_830[FLOAT, 112x336x1x1]
%onnx::Conv_833[FLOAT, 336x112x1x1]
%onnx::Conv_836[FLOAT, 336x1x3x3]
%onnx::Conv_839[FLOAT, 112x336x1x1]
%onnx::Conv_842[FLOAT, 336x112x1x1]
%onnx::Conv_845[FLOAT, 336x1x3x3]
%onnx::Conv_848[FLOAT, 112x336x1x1]
%onnx::Conv_851[FLOAT, 184x112x1x1]
%onnx::Conv_852[FLOAT, 184]
%onnx::Conv_854[FLOAT, 184x92x1x1]
%onnx::Conv_857[FLOAT, 184x1x5x5]
%onnx::Conv_860[FLOAT, 184x92x1x1]
%onnx::Conv_863[FLOAT, 184x92x1x1]
%onnx::Conv_866[FLOAT, 184x1x3x3]
%onnx::Conv_869[FLOAT, 184x92x1x1]
%onnx::Conv_872[FLOAT, 1104x184x1x1]
%onnx::Conv_873[FLOAT, 1104]
%onnx::Conv_875[FLOAT, 1104x1x5x5]
%onnx::Conv_878[FLOAT, 184x1104x1x1]
%onnx::Conv_881[FLOAT, 184x92x1x1]
%onnx::Conv_884[FLOAT, 184x1x3x3]
%onnx::Conv_887[FLOAT, 352x92x1x1]
%onnx::Conv_888[FLOAT, 352]
%onnx::Conv_890[FLOAT, 1504x352x1x1]
%onnx::Conv_891[FLOAT, 1504]
) {
%onnx::Conv_885 = Identity(%onnx::Conv_852)
%onnx::Conv_882 = Identity(%onnx::Conv_852)
%onnx::Conv_879 = Identity(%onnx::Conv_852)
%onnx::Conv_876 = Identity(%onnx::Conv_873)
%onnx::Conv_870 = Identity(%onnx::Conv_852)
%onnx::Conv_867 = Identity(%onnx::Conv_852)
%onnx::Conv_864 = Identity(%onnx::Conv_852)
%onnx::Conv_861 = Identity(%onnx::Conv_852)
%onnx::Conv_858 = Identity(%onnx::Conv_852)
%onnx::Conv_855 = Identity(%onnx::Conv_852)
%onnx::Conv_849 = Identity(%onnx::Conv_822)
%onnx::Conv_846 = Identity(%onnx::Conv_825)
%onnx::Conv_843 = Identity(%onnx::Conv_825)
%onnx::Conv_840 = Identity(%onnx::Conv_822)
%onnx::Conv_837 = Identity(%onnx::Conv_825)
%onnx::Conv_834 = Identity(%onnx::Conv_825)
%onnx::Conv_831 = Identity(%onnx::Conv_822)
%onnx::Conv_828 = Identity(%onnx::Conv_825)
%onnx::Conv_819 = Identity(%onnx::Conv_795)
%onnx::Conv_816 = Identity(%onnx::Conv_795)
%onnx::Conv_813 = Identity(%onnx::Conv_795)
%onnx::Conv_810 = Identity(%onnx::Conv_795)
%onnx::Conv_807 = Identity(%onnx::Conv_795)
%onnx::Conv_804 = Identity(%onnx::Conv_795)
%onnx::Conv_801 = Identity(%onnx::Conv_795)
%onnx::Conv_798 = Identity(%onnx::Conv_795)
%onnx::Conv_792 = Identity(%onnx::Conv_789)
%onnx::Conv_786 = Identity(%onnx::Conv_759)
%onnx::Conv_783 = Identity(%onnx::Conv_780)
%onnx::Conv_777 = Identity(%onnx::Conv_759)
%onnx::Conv_774 = Identity(%onnx::Conv_759)
%onnx::Conv_771 = Identity(%onnx::Conv_759)
%onnx::Conv_768 = Identity(%onnx::Conv_759)
%onnx::Conv_765 = Identity(%onnx::Conv_759)
%onnx::Conv_762 = Identity(%onnx::Conv_759)
%onnx::Conv_756 = Identity(%onnx::Conv_753)
%onnx::Conv_750 = Identity(%onnx::Conv_723)
%onnx::Conv_747 = Identity(%onnx::Conv_723)
%onnx::Conv_744 = Identity(%onnx::Conv_723)
%onnx::Conv_741 = Identity(%onnx::Conv_723)
%onnx::Conv_738 = Identity(%onnx::Conv_723)
%onnx::Conv_735 = Identity(%onnx::Conv_723)
%onnx::Conv_732 = Identity(%onnx::Conv_723)
%onnx::Conv_729 = Identity(%onnx::Conv_726)
%onnx::Conv_720 = Identity(%onnx::Conv_705)
%onnx::Conv_717 = Identity(%onnx::Conv_705)
%onnx::Conv_714 = Identity(%onnx::Conv_705)
%onnx::Conv_711 = Identity(%onnx::Conv_708)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_704, %onnx::Conv_705)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_803, %onnx::Conv_804)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_806, %onnx::Conv_807)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_809, %onnx::Conv_810)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_812, %onnx::Conv_813)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_815, %onnx::Conv_816)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_818, %onnx::Conv_819)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_821, %onnx::Conv_822)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_824, %onnx::Conv_825)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_827, %onnx::Conv_828)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_830, %onnx::Conv_831)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_833, %onnx::Conv_834)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_836, %onnx::Conv_837)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_839, %onnx::Conv_840)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_842, %onnx::Conv_843)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_845, %onnx::Conv_846)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_848, %onnx::Conv_849)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_851, %onnx::Conv_852)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_854, %onnx::Conv_855)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_output_0 = Reshape(%/cells.18/nl/Relu_output_0, %/cells.18/shuffle/Constant_output_0)
%/cells.18/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.18/shuffle/Reshape_output_0)
%/cells.18/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_1_output_0 = Reshape(%/cells.18/shuffle/Transpose_output_0, %/cells.18/shuffle/Constant_1_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/shuffle/Reshape_1_output_0, %onnx::Conv_857, %onnx::Conv_858)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_860, %onnx::Conv_861)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_863, %onnx::Conv_864)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_866, %onnx::Conv_867)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_869, %onnx::Conv_870)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_872, %onnx::Conv_873)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_875, %onnx::Conv_876)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_878, %onnx::Conv_879)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_881, %onnx::Conv_882)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_884, %onnx::Conv_885)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_887, %onnx::Conv_888)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_890, %onnx::Conv_891)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%702 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %702
}
|
val_accuracy
| 0
| 63,470,080
| 1,592,516
|
{'zcp_synflow': 73.9089494920535, 'zcp_zen': 64.18143463134766, 'zcp_epe_nas': 8.917556680646907, 'zcp_fisher': 0.1647902876138687, 'zcp_flops': 63470080.0, 'zcp_grad_norm': 24.628204345703125, 'zcp_grasp': -0.6651763916015625, 'zcp_jacov': -16.062429067386024, 'zcp_l2_norm': 568.1152954101562, 'zcp_nwot': 211.79401294973104, 'zcp_params': 1592516.0, 'zcp_plain': -0.0006641061627306044, 'zcp_snip': 45.354312896728516, 'lat_1080ti_1': 0.6456088834190387, 'lat_1080ti_32': 0.6393953080006235, 'lat_1080ti_64': 0.48334700369806805, 'lat_2080ti_1': 0.702187704035427, 'lat_2080ti_32': 0.6586048571817029, 'lat_2080ti_64': 0.5214851142957196, 'lat_essential_ph_1': 0.1509433962264151, 'lat_eyeriss': 0.36998698036051547, 'lat_fpga': 0.3966890628246867, 'lat_gold_6226': 0.26850189892218723, 'lat_gold_6240': 0.4384205285112476, 'lat_pixel2': 0.32608695652173914, 'lat_pixel3': 0.4004983678472757, 'lat_raspi4': 0.41370788155004784, 'lat_samsung_a50': 0.16842105263157894, 'lat_samsung_s7': 0.44881889763779526, 'lat_silver_4114': 0.46358311128905705, 'lat_silver_4210r': 0.46766948448392964, 'lat_titan_rtx_1': 0.667876388143536, 'lat_titan_rtx_32': 0.6526058984603675, 'lat_titan_rtx_64': 0.5503115133557872, 'lat_titanx_1': 0.34723877162921873, 'lat_titanx_32': 0.5842901889018542, 'lat_titanx_64': 0.4655184944010841, 'lat_titanxp_1': 0.6218144934336531, 'lat_titanxp_32': 0.6320042923028621, 'lat_titanxp_64': 0.5078314872814985}
| |
FBNet_3805
|
FBNet
|
3805
|
3805
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_623[FLOAT, 16x3x3x3]
%onnx::Conv_624[FLOAT, 16]
%onnx::Conv_626[FLOAT, 16x8x1x1]
%onnx::Conv_629[FLOAT, 16x1x5x5]
%onnx::Conv_632[FLOAT, 16x8x1x1]
%onnx::Conv_635[FLOAT, 16x8x1x1]
%onnx::Conv_638[FLOAT, 16x1x5x5]
%onnx::Conv_641[FLOAT, 24x8x1x1]
%onnx::Conv_642[FLOAT, 24]
%onnx::Conv_644[FLOAT, 24x24x1x1]
%onnx::Conv_647[FLOAT, 24x1x3x3]
%onnx::Conv_650[FLOAT, 24x24x1x1]
%onnx::Conv_653[FLOAT, 72x24x1x1]
%onnx::Conv_654[FLOAT, 72]
%onnx::Conv_656[FLOAT, 72x1x3x3]
%onnx::Conv_659[FLOAT, 24x72x1x1]
%onnx::Conv_662[FLOAT, 24x12x1x1]
%onnx::Conv_665[FLOAT, 24x1x3x3]
%onnx::Conv_668[FLOAT, 24x12x1x1]
%onnx::Conv_671[FLOAT, 144x24x1x1]
%onnx::Conv_672[FLOAT, 144]
%onnx::Conv_674[FLOAT, 144x1x5x5]
%onnx::Conv_677[FLOAT, 32x144x1x1]
%onnx::Conv_678[FLOAT, 32]
%onnx::Conv_680[FLOAT, 96x32x1x1]
%onnx::Conv_681[FLOAT, 96]
%onnx::Conv_683[FLOAT, 96x1x5x5]
%onnx::Conv_686[FLOAT, 32x96x1x1]
%onnx::Conv_689[FLOAT, 192x32x1x1]
%onnx::Conv_690[FLOAT, 192]
%onnx::Conv_692[FLOAT, 192x1x5x5]
%onnx::Conv_695[FLOAT, 32x192x1x1]
%onnx::Conv_698[FLOAT, 32x32x1x1]
%onnx::Conv_701[FLOAT, 32x1x3x3]
%onnx::Conv_704[FLOAT, 64x32x1x1]
%onnx::Conv_705[FLOAT, 64]
%onnx::Conv_707[FLOAT, 64x64x1x1]
%onnx::Conv_710[FLOAT, 64x1x5x5]
%onnx::Conv_713[FLOAT, 64x64x1x1]
%onnx::Conv_716[FLOAT, 384x64x1x1]
%onnx::Conv_717[FLOAT, 384]
%onnx::Conv_719[FLOAT, 384x1x5x5]
%onnx::Conv_722[FLOAT, 64x384x1x1]
%onnx::Conv_725[FLOAT, 64x64x1x1]
%onnx::Conv_728[FLOAT, 64x1x5x5]
%onnx::Conv_731[FLOAT, 64x64x1x1]
%onnx::Conv_734[FLOAT, 384x64x1x1]
%onnx::Conv_737[FLOAT, 384x1x5x5]
%onnx::Conv_740[FLOAT, 112x384x1x1]
%onnx::Conv_741[FLOAT, 112]
%onnx::Conv_743[FLOAT, 112x56x1x1]
%onnx::Conv_746[FLOAT, 112x1x3x3]
%onnx::Conv_749[FLOAT, 112x56x1x1]
%onnx::Conv_752[FLOAT, 336x112x1x1]
%onnx::Conv_753[FLOAT, 336]
%onnx::Conv_755[FLOAT, 336x1x3x3]
%onnx::Conv_758[FLOAT, 112x336x1x1]
%onnx::Conv_761[FLOAT, 184x112x1x1]
%onnx::Conv_762[FLOAT, 184]
%onnx::Conv_764[FLOAT, 184x184x1x1]
%onnx::Conv_767[FLOAT, 184x1x5x5]
%onnx::Conv_770[FLOAT, 184x184x1x1]
%onnx::Conv_773[FLOAT, 1104x184x1x1]
%onnx::Conv_774[FLOAT, 1104]
%onnx::Conv_776[FLOAT, 1104x1x3x3]
%onnx::Conv_779[FLOAT, 184x1104x1x1]
%onnx::Conv_782[FLOAT, 1104x184x1x1]
%onnx::Conv_785[FLOAT, 1104x1x5x5]
%onnx::Conv_788[FLOAT, 184x1104x1x1]
%onnx::Conv_791[FLOAT, 184x184x1x1]
%onnx::Conv_794[FLOAT, 184x1x3x3]
%onnx::Conv_797[FLOAT, 352x184x1x1]
%onnx::Conv_798[FLOAT, 352]
%onnx::Conv_800[FLOAT, 1504x352x1x1]
%onnx::Conv_801[FLOAT, 1504]
) {
%onnx::Conv_795 = Identity(%onnx::Conv_762)
%onnx::Conv_792 = Identity(%onnx::Conv_762)
%onnx::Conv_789 = Identity(%onnx::Conv_762)
%onnx::Conv_786 = Identity(%onnx::Conv_774)
%onnx::Conv_783 = Identity(%onnx::Conv_774)
%onnx::Conv_780 = Identity(%onnx::Conv_762)
%onnx::Conv_777 = Identity(%onnx::Conv_774)
%onnx::Conv_771 = Identity(%onnx::Conv_762)
%onnx::Conv_768 = Identity(%onnx::Conv_762)
%onnx::Conv_765 = Identity(%onnx::Conv_762)
%onnx::Conv_759 = Identity(%onnx::Conv_741)
%onnx::Conv_756 = Identity(%onnx::Conv_753)
%onnx::Conv_750 = Identity(%onnx::Conv_741)
%onnx::Conv_747 = Identity(%onnx::Conv_741)
%onnx::Conv_744 = Identity(%onnx::Conv_741)
%onnx::Conv_738 = Identity(%onnx::Conv_717)
%onnx::Conv_735 = Identity(%onnx::Conv_717)
%onnx::Conv_732 = Identity(%onnx::Conv_705)
%onnx::Conv_729 = Identity(%onnx::Conv_705)
%onnx::Conv_726 = Identity(%onnx::Conv_705)
%onnx::Conv_723 = Identity(%onnx::Conv_705)
%onnx::Conv_720 = Identity(%onnx::Conv_717)
%onnx::Conv_714 = Identity(%onnx::Conv_705)
%onnx::Conv_711 = Identity(%onnx::Conv_705)
%onnx::Conv_708 = Identity(%onnx::Conv_705)
%onnx::Conv_702 = Identity(%onnx::Conv_678)
%onnx::Conv_699 = Identity(%onnx::Conv_678)
%onnx::Conv_696 = Identity(%onnx::Conv_678)
%onnx::Conv_693 = Identity(%onnx::Conv_690)
%onnx::Conv_687 = Identity(%onnx::Conv_678)
%onnx::Conv_684 = Identity(%onnx::Conv_681)
%onnx::Conv_675 = Identity(%onnx::Conv_672)
%onnx::Conv_669 = Identity(%onnx::Conv_642)
%onnx::Conv_666 = Identity(%onnx::Conv_642)
%onnx::Conv_663 = Identity(%onnx::Conv_642)
%onnx::Conv_660 = Identity(%onnx::Conv_642)
%onnx::Conv_657 = Identity(%onnx::Conv_654)
%onnx::Conv_651 = Identity(%onnx::Conv_642)
%onnx::Conv_648 = Identity(%onnx::Conv_642)
%onnx::Conv_645 = Identity(%onnx::Conv_642)
%onnx::Conv_639 = Identity(%onnx::Conv_624)
%onnx::Conv_636 = Identity(%onnx::Conv_624)
%onnx::Conv_633 = Identity(%onnx::Conv_624)
%onnx::Conv_630 = Identity(%onnx::Conv_624)
%onnx::Conv_627 = Identity(%onnx::Conv_624)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_623, %onnx::Conv_624)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%621 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %621
}
|
val_accuracy
| 0
| 64,772,352
| 2,044,236
|
{'zcp_synflow': 75.99663290327742, 'zcp_zen': 65.6636962890625, 'zcp_epe_nas': 16.120919608431787, 'zcp_fisher': 0.11191096901893616, 'zcp_flops': 64772352.0, 'zcp_grad_norm': 22.671993255615234, 'zcp_grasp': -0.028392791748046875, 'zcp_jacov': -16.054835630710375, 'zcp_l2_norm': 614.120849609375, 'zcp_nwot': 210.69115910326806, 'zcp_params': 2044236.0, 'zcp_plain': -0.008627538569271564, 'zcp_snip': 35.98661422729492, 'lat_1080ti_1': 0.5394920494003461, 'lat_1080ti_32': 0.3943345943893492, 'lat_1080ti_64': 0.36357157590107136, 'lat_2080ti_1': 0.5494861937631289, 'lat_2080ti_32': 0.45155945189912033, 'lat_2080ti_64': 0.3524406228570524, 'lat_essential_ph_1': 0.16981132075471697, 'lat_eyeriss': 0.4319127867529752, 'lat_fpga': 0.4228124165091677, 'lat_gold_6226': 0.38609207241056925, 'lat_gold_6240': 0.4902936786949775, 'lat_pixel2': 0.43478260869565216, 'lat_pixel3': 0.41459628812675786, 'lat_raspi4': 0.4553115521595239, 'lat_samsung_a50': 0.18947368421052632, 'lat_samsung_s7': 0.5354330708661418, 'lat_silver_4114': 0.6104174772330427, 'lat_silver_4210r': 0.5198783781307196, 'lat_titan_rtx_1': 0.5148312980777615, 'lat_titan_rtx_32': 0.4394507340798605, 'lat_titan_rtx_64': 0.36676812718372626, 'lat_titanx_1': 0.2751516349721012, 'lat_titanx_32': 0.37838125743347173, 'lat_titanx_64': 0.38209887574902907, 'lat_titanxp_1': 0.48196573182648467, 'lat_titanxp_32': 0.4047433968843784, 'lat_titanxp_64': 0.35649707656431195}
| |
FBNet_630
|
FBNet
|
630
|
630
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_686[FLOAT, 16x3x3x3]
%onnx::Conv_687[FLOAT, 16]
%onnx::Conv_689[FLOAT, 96x16x1x1]
%onnx::Conv_690[FLOAT, 96]
%onnx::Conv_692[FLOAT, 96x1x5x5]
%onnx::Conv_695[FLOAT, 16x96x1x1]
%onnx::Conv_698[FLOAT, 16x16x1x1]
%onnx::Conv_701[FLOAT, 16x1x3x3]
%onnx::Conv_704[FLOAT, 24x16x1x1]
%onnx::Conv_705[FLOAT, 24]
%onnx::Conv_707[FLOAT, 72x24x1x1]
%onnx::Conv_708[FLOAT, 72]
%onnx::Conv_710[FLOAT, 72x1x5x5]
%onnx::Conv_713[FLOAT, 24x72x1x1]
%onnx::Conv_716[FLOAT, 144x24x1x1]
%onnx::Conv_717[FLOAT, 144]
%onnx::Conv_719[FLOAT, 144x1x5x5]
%onnx::Conv_722[FLOAT, 24x144x1x1]
%onnx::Conv_725[FLOAT, 24x12x1x1]
%onnx::Conv_728[FLOAT, 24x1x5x5]
%onnx::Conv_731[FLOAT, 32x12x1x1]
%onnx::Conv_732[FLOAT, 32]
%onnx::Conv_734[FLOAT, 192x32x1x1]
%onnx::Conv_735[FLOAT, 192]
%onnx::Conv_737[FLOAT, 192x1x3x3]
%onnx::Conv_740[FLOAT, 32x192x1x1]
%onnx::Conv_743[FLOAT, 32x32x1x1]
%onnx::Conv_746[FLOAT, 32x1x5x5]
%onnx::Conv_749[FLOAT, 32x32x1x1]
%onnx::Conv_752[FLOAT, 96x32x1x1]
%onnx::Conv_755[FLOAT, 96x1x3x3]
%onnx::Conv_758[FLOAT, 32x96x1x1]
%onnx::Conv_761[FLOAT, 32x32x1x1]
%onnx::Conv_764[FLOAT, 32x1x5x5]
%onnx::Conv_767[FLOAT, 64x32x1x1]
%onnx::Conv_768[FLOAT, 64]
%onnx::Conv_770[FLOAT, 64x64x1x1]
%onnx::Conv_773[FLOAT, 64x1x5x5]
%onnx::Conv_776[FLOAT, 64x64x1x1]
%onnx::Conv_779[FLOAT, 64x64x1x1]
%onnx::Conv_782[FLOAT, 64x1x3x3]
%onnx::Conv_785[FLOAT, 64x64x1x1]
%onnx::Conv_788[FLOAT, 64x32x1x1]
%onnx::Conv_791[FLOAT, 64x1x3x3]
%onnx::Conv_794[FLOAT, 64x32x1x1]
%onnx::Conv_797[FLOAT, 384x64x1x1]
%onnx::Conv_798[FLOAT, 384]
%onnx::Conv_800[FLOAT, 384x1x3x3]
%onnx::Conv_803[FLOAT, 112x384x1x1]
%onnx::Conv_804[FLOAT, 112]
%onnx::Conv_806[FLOAT, 336x112x1x1]
%onnx::Conv_807[FLOAT, 336]
%onnx::Conv_809[FLOAT, 336x1x5x5]
%onnx::Conv_812[FLOAT, 112x336x1x1]
%onnx::Conv_815[FLOAT, 112x56x1x1]
%onnx::Conv_818[FLOAT, 112x1x5x5]
%onnx::Conv_821[FLOAT, 112x56x1x1]
%onnx::Conv_824[FLOAT, 336x112x1x1]
%onnx::Conv_827[FLOAT, 336x1x3x3]
%onnx::Conv_830[FLOAT, 112x336x1x1]
%onnx::Conv_833[FLOAT, 112x112x1x1]
%onnx::Conv_836[FLOAT, 112x1x3x3]
%onnx::Conv_839[FLOAT, 184x112x1x1]
%onnx::Conv_840[FLOAT, 184]
%onnx::Conv_842[FLOAT, 1104x184x1x1]
%onnx::Conv_843[FLOAT, 1104]
%onnx::Conv_845[FLOAT, 1104x1x5x5]
%onnx::Conv_848[FLOAT, 184x1104x1x1]
%onnx::Conv_851[FLOAT, 184x92x1x1]
%onnx::Conv_854[FLOAT, 184x1x3x3]
%onnx::Conv_857[FLOAT, 184x92x1x1]
%onnx::Conv_860[FLOAT, 1104x184x1x1]
%onnx::Conv_863[FLOAT, 1104x1x3x3]
%onnx::Conv_866[FLOAT, 184x1104x1x1]
%onnx::Conv_869[FLOAT, 184x92x1x1]
%onnx::Conv_872[FLOAT, 184x1x3x3]
%onnx::Conv_875[FLOAT, 352x92x1x1]
%onnx::Conv_876[FLOAT, 352]
%onnx::Conv_878[FLOAT, 1504x352x1x1]
%onnx::Conv_879[FLOAT, 1504]
) {
%onnx::Conv_873 = Identity(%onnx::Conv_840)
%onnx::Conv_870 = Identity(%onnx::Conv_840)
%onnx::Conv_867 = Identity(%onnx::Conv_840)
%onnx::Conv_864 = Identity(%onnx::Conv_843)
%onnx::Conv_861 = Identity(%onnx::Conv_843)
%onnx::Conv_858 = Identity(%onnx::Conv_840)
%onnx::Conv_855 = Identity(%onnx::Conv_840)
%onnx::Conv_852 = Identity(%onnx::Conv_840)
%onnx::Conv_849 = Identity(%onnx::Conv_840)
%onnx::Conv_846 = Identity(%onnx::Conv_843)
%onnx::Conv_837 = Identity(%onnx::Conv_804)
%onnx::Conv_834 = Identity(%onnx::Conv_804)
%onnx::Conv_831 = Identity(%onnx::Conv_804)
%onnx::Conv_828 = Identity(%onnx::Conv_807)
%onnx::Conv_825 = Identity(%onnx::Conv_807)
%onnx::Conv_822 = Identity(%onnx::Conv_804)
%onnx::Conv_819 = Identity(%onnx::Conv_804)
%onnx::Conv_816 = Identity(%onnx::Conv_804)
%onnx::Conv_813 = Identity(%onnx::Conv_804)
%onnx::Conv_810 = Identity(%onnx::Conv_807)
%onnx::Conv_801 = Identity(%onnx::Conv_798)
%onnx::Conv_795 = Identity(%onnx::Conv_768)
%onnx::Conv_792 = Identity(%onnx::Conv_768)
%onnx::Conv_789 = Identity(%onnx::Conv_768)
%onnx::Conv_786 = Identity(%onnx::Conv_768)
%onnx::Conv_783 = Identity(%onnx::Conv_768)
%onnx::Conv_780 = Identity(%onnx::Conv_768)
%onnx::Conv_777 = Identity(%onnx::Conv_768)
%onnx::Conv_774 = Identity(%onnx::Conv_768)
%onnx::Conv_771 = Identity(%onnx::Conv_768)
%onnx::Conv_765 = Identity(%onnx::Conv_732)
%onnx::Conv_762 = Identity(%onnx::Conv_732)
%onnx::Conv_759 = Identity(%onnx::Conv_732)
%onnx::Conv_756 = Identity(%onnx::Conv_690)
%onnx::Conv_753 = Identity(%onnx::Conv_690)
%onnx::Conv_750 = Identity(%onnx::Conv_732)
%onnx::Conv_747 = Identity(%onnx::Conv_732)
%onnx::Conv_744 = Identity(%onnx::Conv_732)
%onnx::Conv_741 = Identity(%onnx::Conv_732)
%onnx::Conv_738 = Identity(%onnx::Conv_735)
%onnx::Conv_729 = Identity(%onnx::Conv_705)
%onnx::Conv_726 = Identity(%onnx::Conv_705)
%onnx::Conv_723 = Identity(%onnx::Conv_705)
%onnx::Conv_720 = Identity(%onnx::Conv_717)
%onnx::Conv_714 = Identity(%onnx::Conv_705)
%onnx::Conv_711 = Identity(%onnx::Conv_708)
%onnx::Conv_702 = Identity(%onnx::Conv_687)
%onnx::Conv_699 = Identity(%onnx::Conv_687)
%onnx::Conv_696 = Identity(%onnx::Conv_687)
%onnx::Conv_693 = Identity(%onnx::Conv_690)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_686, %onnx::Conv_687)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_803, %onnx::Conv_804)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_806, %onnx::Conv_807)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_809, %onnx::Conv_810)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_812, %onnx::Conv_813)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_815, %onnx::Conv_816)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_818, %onnx::Conv_819)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_821, %onnx::Conv_822)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_824, %onnx::Conv_825)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_827, %onnx::Conv_828)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_830, %onnx::Conv_831)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_833, %onnx::Conv_834)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_836, %onnx::Conv_837)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_839, %onnx::Conv_840)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_842, %onnx::Conv_843)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_845, %onnx::Conv_846)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_848, %onnx::Conv_849)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_851, %onnx::Conv_852)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_854, %onnx::Conv_855)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_857, %onnx::Conv_858)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_860, %onnx::Conv_861)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_863, %onnx::Conv_864)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_866, %onnx::Conv_867)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_869, %onnx::Conv_870)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_872, %onnx::Conv_873)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_875, %onnx::Conv_876)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_878, %onnx::Conv_879)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%684 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %684
}
|
val_accuracy
| 0
| 75,588,480
| 1,999,732
|
{'zcp_synflow': 79.72276324966414, 'zcp_zen': 70.59368133544922, 'zcp_epe_nas': 21.467019266403636, 'zcp_fisher': 0.22774608433246613, 'zcp_flops': 75588480.0, 'zcp_grad_norm': 29.253814697265625, 'zcp_grasp': -0.1461944580078125, 'zcp_jacov': -16.07644902456935, 'zcp_l2_norm': 641.7670288085938, 'zcp_nwot': 214.22117472564213, 'zcp_params': 1999732.0, 'zcp_plain': -0.0006176821771077812, 'zcp_snip': 50.00994110107422, 'lat_1080ti_1': 0.6320446548297504, 'lat_1080ti_32': 0.7430505377117663, 'lat_1080ti_64': 0.6390208060730067, 'lat_2080ti_1': 0.720472451499225, 'lat_2080ti_32': 0.732438372284741, 'lat_2080ti_64': 0.6337911646882272, 'lat_essential_ph_1': 0.32075471698113206, 'lat_eyeriss': 0.574065244897585, 'lat_fpga': 0.5652836461868809, 'lat_gold_6226': 0.3866623093197568, 'lat_gold_6240': 0.5734131617786282, 'lat_pixel2': 0.3695652173913043, 'lat_pixel3': 0.5838156610818735, 'lat_raspi4': 0.5907597070162165, 'lat_samsung_a50': 0.23157894736842105, 'lat_samsung_s7': 0.1968503937007874, 'lat_silver_4114': 0.640024297077483, 'lat_silver_4210r': 0.6332662069424203, 'lat_titan_rtx_1': 0.6810518141065961, 'lat_titan_rtx_32': 0.7296595183983496, 'lat_titan_rtx_64': 0.6822580708098454, 'lat_titanx_1': 0.37234601209292534, 'lat_titanx_32': 0.7136346019408643, 'lat_titanx_64': 0.6034973819996521, 'lat_titanxp_1': 0.638215591885592, 'lat_titanxp_32': 0.7392322846480491, 'lat_titanxp_64': 0.6423458267051984}
| |
FBNet_3230
|
FBNet
|
3230
|
3230
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_634[FLOAT, 16x3x3x3]
%onnx::Conv_635[FLOAT, 16]
%onnx::Conv_637[FLOAT, 96x16x1x1]
%onnx::Conv_638[FLOAT, 96]
%onnx::Conv_640[FLOAT, 96x1x5x5]
%onnx::Conv_643[FLOAT, 24x96x1x1]
%onnx::Conv_644[FLOAT, 24]
%onnx::Conv_646[FLOAT, 24x24x1x1]
%onnx::Conv_649[FLOAT, 24x1x5x5]
%onnx::Conv_652[FLOAT, 24x24x1x1]
%onnx::Conv_655[FLOAT, 24x24x1x1]
%onnx::Conv_658[FLOAT, 24x1x5x5]
%onnx::Conv_661[FLOAT, 24x24x1x1]
%onnx::Conv_664[FLOAT, 144x24x1x1]
%onnx::Conv_665[FLOAT, 144]
%onnx::Conv_667[FLOAT, 144x1x3x3]
%onnx::Conv_670[FLOAT, 32x144x1x1]
%onnx::Conv_671[FLOAT, 32]
%onnx::Conv_673[FLOAT, 32x16x1x1]
%onnx::Conv_676[FLOAT, 32x1x3x3]
%onnx::Conv_679[FLOAT, 32x16x1x1]
%onnx::Conv_682[FLOAT, 32x16x1x1]
%onnx::Conv_685[FLOAT, 32x1x5x5]
%onnx::Conv_688[FLOAT, 32x16x1x1]
%onnx::Conv_691[FLOAT, 32x16x1x1]
%onnx::Conv_694[FLOAT, 32x1x5x5]
%onnx::Conv_697[FLOAT, 32x16x1x1]
%onnx::Conv_700[FLOAT, 64x32x1x1]
%onnx::Conv_701[FLOAT, 64]
%onnx::Conv_703[FLOAT, 64x32x1x1]
%onnx::Conv_706[FLOAT, 64x1x5x5]
%onnx::Conv_709[FLOAT, 64x32x1x1]
%onnx::Conv_712[FLOAT, 64x64x1x1]
%onnx::Conv_715[FLOAT, 64x1x5x5]
%onnx::Conv_718[FLOAT, 64x64x1x1]
%onnx::Conv_721[FLOAT, 64x32x1x1]
%onnx::Conv_724[FLOAT, 64x1x3x3]
%onnx::Conv_727[FLOAT, 64x32x1x1]
%onnx::Conv_730[FLOAT, 112x64x1x1]
%onnx::Conv_731[FLOAT, 112]
%onnx::Conv_733[FLOAT, 112x56x1x1]
%onnx::Conv_736[FLOAT, 112x1x3x3]
%onnx::Conv_739[FLOAT, 112x56x1x1]
%onnx::Conv_742[FLOAT, 672x112x1x1]
%onnx::Conv_743[FLOAT, 672]
%onnx::Conv_745[FLOAT, 672x1x5x5]
%onnx::Conv_748[FLOAT, 112x672x1x1]
%onnx::Conv_751[FLOAT, 672x112x1x1]
%onnx::Conv_754[FLOAT, 672x1x3x3]
%onnx::Conv_757[FLOAT, 184x672x1x1]
%onnx::Conv_758[FLOAT, 184]
%onnx::Conv_760[FLOAT, 1104x184x1x1]
%onnx::Conv_761[FLOAT, 1104]
%onnx::Conv_763[FLOAT, 1104x1x3x3]
%onnx::Conv_766[FLOAT, 184x1104x1x1]
%onnx::Conv_769[FLOAT, 552x184x1x1]
%onnx::Conv_770[FLOAT, 552]
%onnx::Conv_772[FLOAT, 552x1x3x3]
%onnx::Conv_775[FLOAT, 184x552x1x1]
%onnx::Conv_778[FLOAT, 184x92x1x1]
%onnx::Conv_781[FLOAT, 184x1x3x3]
%onnx::Conv_784[FLOAT, 184x92x1x1]
%onnx::Conv_787[FLOAT, 184x184x1x1]
%onnx::Conv_790[FLOAT, 184x1x3x3]
%onnx::Conv_793[FLOAT, 352x184x1x1]
%onnx::Conv_794[FLOAT, 352]
%onnx::Conv_796[FLOAT, 1504x352x1x1]
%onnx::Conv_797[FLOAT, 1504]
) {
%onnx::Conv_791 = Identity(%onnx::Conv_758)
%onnx::Conv_788 = Identity(%onnx::Conv_758)
%onnx::Conv_785 = Identity(%onnx::Conv_758)
%onnx::Conv_782 = Identity(%onnx::Conv_758)
%onnx::Conv_779 = Identity(%onnx::Conv_758)
%onnx::Conv_776 = Identity(%onnx::Conv_758)
%onnx::Conv_773 = Identity(%onnx::Conv_770)
%onnx::Conv_767 = Identity(%onnx::Conv_758)
%onnx::Conv_764 = Identity(%onnx::Conv_761)
%onnx::Conv_755 = Identity(%onnx::Conv_743)
%onnx::Conv_752 = Identity(%onnx::Conv_743)
%onnx::Conv_749 = Identity(%onnx::Conv_731)
%onnx::Conv_746 = Identity(%onnx::Conv_743)
%onnx::Conv_740 = Identity(%onnx::Conv_731)
%onnx::Conv_737 = Identity(%onnx::Conv_731)
%onnx::Conv_734 = Identity(%onnx::Conv_731)
%onnx::Conv_728 = Identity(%onnx::Conv_701)
%onnx::Conv_725 = Identity(%onnx::Conv_701)
%onnx::Conv_722 = Identity(%onnx::Conv_701)
%onnx::Conv_719 = Identity(%onnx::Conv_701)
%onnx::Conv_716 = Identity(%onnx::Conv_701)
%onnx::Conv_713 = Identity(%onnx::Conv_701)
%onnx::Conv_710 = Identity(%onnx::Conv_701)
%onnx::Conv_707 = Identity(%onnx::Conv_701)
%onnx::Conv_704 = Identity(%onnx::Conv_701)
%onnx::Conv_698 = Identity(%onnx::Conv_671)
%onnx::Conv_695 = Identity(%onnx::Conv_671)
%onnx::Conv_692 = Identity(%onnx::Conv_671)
%onnx::Conv_689 = Identity(%onnx::Conv_671)
%onnx::Conv_686 = Identity(%onnx::Conv_671)
%onnx::Conv_683 = Identity(%onnx::Conv_671)
%onnx::Conv_680 = Identity(%onnx::Conv_671)
%onnx::Conv_677 = Identity(%onnx::Conv_671)
%onnx::Conv_674 = Identity(%onnx::Conv_671)
%onnx::Conv_668 = Identity(%onnx::Conv_665)
%onnx::Conv_662 = Identity(%onnx::Conv_644)
%onnx::Conv_659 = Identity(%onnx::Conv_644)
%onnx::Conv_656 = Identity(%onnx::Conv_644)
%onnx::Conv_653 = Identity(%onnx::Conv_644)
%onnx::Conv_650 = Identity(%onnx::Conv_644)
%onnx::Conv_647 = Identity(%onnx::Conv_644)
%onnx::Conv_641 = Identity(%onnx::Conv_638)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_634, %onnx::Conv_635)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_637, %onnx::Conv_638)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_640, %onnx::Conv_641)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_643, %onnx::Conv_644)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_646, %onnx::Conv_647)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_649, %onnx::Conv_650)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_652, %onnx::Conv_653)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_655, %onnx::Conv_656)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_658, %onnx::Conv_659)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.8/Add_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.9/relu/Relu_output_0 = Relu(%/cells.9/conv/Conv_output_0)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/relu/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/relu/Relu_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.13/relu/Relu_output_0 = Relu(%/cells.13/conv/Conv_output_0)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/relu/Relu_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/relu/Relu_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%632 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %632
}
|
val_accuracy
| 0
| 60,209,792
| 1,902,748
|
{'zcp_synflow': 67.25310194257851, 'zcp_zen': 58.72089385986328, 'zcp_epe_nas': 8.226296113033234, 'zcp_fisher': 0.07455726712942123, 'zcp_flops': 60209792.0, 'zcp_grad_norm': 18.127277374267578, 'zcp_grasp': -0.045452117919921875, 'zcp_jacov': -16.057864956897742, 'zcp_l2_norm': 541.3313598632812, 'zcp_nwot': 208.3434051571591, 'zcp_params': 1902748.0, 'zcp_plain': 0.00535919051617384, 'zcp_snip': 34.5508918762207, 'lat_1080ti_1': 0.4078364042765235, 'lat_1080ti_32': 0.39405596591577013, 'lat_1080ti_64': 0.32210293769282033, 'lat_2080ti_1': 0.45320630238951826, 'lat_2080ti_32': 0.400402328900585, 'lat_2080ti_64': 0.3427729122699328, 'lat_essential_ph_1': 0.1320754716981132, 'lat_eyeriss': 0.3001577026754442, 'lat_fpga': 0.29744010924311537, 'lat_gold_6226': 0.3010600725197462, 'lat_gold_6240': 0.407054496868152, 'lat_pixel2': 0.2391304347826087, 'lat_pixel3': 0.3247718980575949, 'lat_raspi4': 0.34885831026740904, 'lat_samsung_a50': 0.15789473684210525, 'lat_samsung_s7': 0.12598425196850394, 'lat_silver_4114': 0.4631578608137486, 'lat_silver_4210r': 0.4435248308863563, 'lat_titan_rtx_1': 0.4119000303028058, 'lat_titan_rtx_32': 0.38321599715915033, 'lat_titan_rtx_64': 0.3403197970045696, 'lat_titanx_1': 0.2350624557168213, 'lat_titanx_32': 0.333477682182724, 'lat_titanx_64': 0.3148408290495659, 'lat_titanxp_1': 0.4134779597125548, 'lat_titanxp_32': 0.38507017669458493, 'lat_titanxp_64': 0.3352902721215779}
| |
FBNet_3208
|
FBNet
|
3208
|
3208
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_596[FLOAT, 16x3x3x3]
%onnx::Conv_597[FLOAT, 16]
%onnx::Conv_599[FLOAT, 96x16x1x1]
%onnx::Conv_600[FLOAT, 96]
%onnx::Conv_602[FLOAT, 96x1x5x5]
%onnx::Conv_605[FLOAT, 16x96x1x1]
%onnx::Conv_608[FLOAT, 96x16x1x1]
%onnx::Conv_611[FLOAT, 96x1x5x5]
%onnx::Conv_614[FLOAT, 24x96x1x1]
%onnx::Conv_615[FLOAT, 24]
%onnx::Conv_617[FLOAT, 24x24x1x1]
%onnx::Conv_620[FLOAT, 24x1x3x3]
%onnx::Conv_623[FLOAT, 24x24x1x1]
%onnx::Conv_626[FLOAT, 24x24x1x1]
%onnx::Conv_629[FLOAT, 24x1x3x3]
%onnx::Conv_632[FLOAT, 24x24x1x1]
%onnx::Conv_635[FLOAT, 24x24x1x1]
%onnx::Conv_638[FLOAT, 24x1x5x5]
%onnx::Conv_641[FLOAT, 32x24x1x1]
%onnx::Conv_642[FLOAT, 32]
%onnx::Conv_644[FLOAT, 96x32x1x1]
%onnx::Conv_647[FLOAT, 96x1x3x3]
%onnx::Conv_650[FLOAT, 32x96x1x1]
%onnx::Conv_653[FLOAT, 32x16x1x1]
%onnx::Conv_656[FLOAT, 32x1x3x3]
%onnx::Conv_659[FLOAT, 32x16x1x1]
%onnx::Conv_662[FLOAT, 64x32x1x1]
%onnx::Conv_663[FLOAT, 64]
%onnx::Conv_665[FLOAT, 64x32x1x1]
%onnx::Conv_668[FLOAT, 64x1x3x3]
%onnx::Conv_671[FLOAT, 64x32x1x1]
%onnx::Conv_674[FLOAT, 192x64x1x1]
%onnx::Conv_675[FLOAT, 192]
%onnx::Conv_677[FLOAT, 192x1x5x5]
%onnx::Conv_680[FLOAT, 64x192x1x1]
%onnx::Conv_683[FLOAT, 64x64x1x1]
%onnx::Conv_686[FLOAT, 64x1x5x5]
%onnx::Conv_689[FLOAT, 64x64x1x1]
%onnx::Conv_692[FLOAT, 64x32x1x1]
%onnx::Conv_695[FLOAT, 64x1x3x3]
%onnx::Conv_698[FLOAT, 112x32x1x1]
%onnx::Conv_699[FLOAT, 112]
%onnx::Conv_701[FLOAT, 112x112x1x1]
%onnx::Conv_704[FLOAT, 112x1x5x5]
%onnx::Conv_707[FLOAT, 112x112x1x1]
%onnx::Conv_710[FLOAT, 672x112x1x1]
%onnx::Conv_711[FLOAT, 672]
%onnx::Conv_713[FLOAT, 672x1x5x5]
%onnx::Conv_716[FLOAT, 112x672x1x1]
%onnx::Conv_719[FLOAT, 336x112x1x1]
%onnx::Conv_720[FLOAT, 336]
%onnx::Conv_722[FLOAT, 336x1x3x3]
%onnx::Conv_725[FLOAT, 112x336x1x1]
%onnx::Conv_728[FLOAT, 112x112x1x1]
%onnx::Conv_731[FLOAT, 112x1x5x5]
%onnx::Conv_734[FLOAT, 184x112x1x1]
%onnx::Conv_735[FLOAT, 184]
%onnx::Conv_737[FLOAT, 552x184x1x1]
%onnx::Conv_738[FLOAT, 552]
%onnx::Conv_740[FLOAT, 552x1x3x3]
%onnx::Conv_743[FLOAT, 184x552x1x1]
%onnx::Conv_746[FLOAT, 184x92x1x1]
%onnx::Conv_749[FLOAT, 184x1x5x5]
%onnx::Conv_752[FLOAT, 184x92x1x1]
%onnx::Conv_755[FLOAT, 1104x184x1x1]
%onnx::Conv_756[FLOAT, 1104]
%onnx::Conv_758[FLOAT, 1104x1x3x3]
%onnx::Conv_761[FLOAT, 352x1104x1x1]
%onnx::Conv_762[FLOAT, 352]
%onnx::Conv_764[FLOAT, 1504x352x1x1]
%onnx::Conv_765[FLOAT, 1504]
) {
%onnx::Conv_759 = Identity(%onnx::Conv_756)
%onnx::Conv_753 = Identity(%onnx::Conv_735)
%onnx::Conv_750 = Identity(%onnx::Conv_735)
%onnx::Conv_747 = Identity(%onnx::Conv_735)
%onnx::Conv_744 = Identity(%onnx::Conv_735)
%onnx::Conv_741 = Identity(%onnx::Conv_738)
%onnx::Conv_732 = Identity(%onnx::Conv_699)
%onnx::Conv_729 = Identity(%onnx::Conv_699)
%onnx::Conv_726 = Identity(%onnx::Conv_699)
%onnx::Conv_723 = Identity(%onnx::Conv_720)
%onnx::Conv_717 = Identity(%onnx::Conv_699)
%onnx::Conv_714 = Identity(%onnx::Conv_711)
%onnx::Conv_708 = Identity(%onnx::Conv_699)
%onnx::Conv_705 = Identity(%onnx::Conv_699)
%onnx::Conv_702 = Identity(%onnx::Conv_699)
%onnx::Conv_696 = Identity(%onnx::Conv_663)
%onnx::Conv_693 = Identity(%onnx::Conv_663)
%onnx::Conv_690 = Identity(%onnx::Conv_663)
%onnx::Conv_687 = Identity(%onnx::Conv_663)
%onnx::Conv_684 = Identity(%onnx::Conv_663)
%onnx::Conv_681 = Identity(%onnx::Conv_663)
%onnx::Conv_678 = Identity(%onnx::Conv_675)
%onnx::Conv_672 = Identity(%onnx::Conv_663)
%onnx::Conv_669 = Identity(%onnx::Conv_663)
%onnx::Conv_666 = Identity(%onnx::Conv_663)
%onnx::Conv_660 = Identity(%onnx::Conv_642)
%onnx::Conv_657 = Identity(%onnx::Conv_642)
%onnx::Conv_654 = Identity(%onnx::Conv_642)
%onnx::Conv_651 = Identity(%onnx::Conv_642)
%onnx::Conv_648 = Identity(%onnx::Conv_600)
%onnx::Conv_645 = Identity(%onnx::Conv_600)
%onnx::Conv_639 = Identity(%onnx::Conv_615)
%onnx::Conv_636 = Identity(%onnx::Conv_615)
%onnx::Conv_633 = Identity(%onnx::Conv_615)
%onnx::Conv_630 = Identity(%onnx::Conv_615)
%onnx::Conv_627 = Identity(%onnx::Conv_615)
%onnx::Conv_624 = Identity(%onnx::Conv_615)
%onnx::Conv_621 = Identity(%onnx::Conv_615)
%onnx::Conv_618 = Identity(%onnx::Conv_615)
%onnx::Conv_612 = Identity(%onnx::Conv_600)
%onnx::Conv_609 = Identity(%onnx::Conv_600)
%onnx::Conv_606 = Identity(%onnx::Conv_597)
%onnx::Conv_603 = Identity(%onnx::Conv_600)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_596, %onnx::Conv_597)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_599, %onnx::Conv_600)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_602, %onnx::Conv_603)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_605, %onnx::Conv_606)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_608, %onnx::Conv_609)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_611, %onnx::Conv_612)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_614, %onnx::Conv_615)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_617, %onnx::Conv_618)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_620, %onnx::Conv_621)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_623, %onnx::Conv_624)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.8/Add_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.9/relu/Relu_output_0 = Relu(%/cells.9/conv/Conv_output_0)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/relu/Relu_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/relu/Relu_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%594 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %594
}
|
val_accuracy
| 0
| 65,121,920
| 1,936,860
|
{'zcp_synflow': 71.88004999278044, 'zcp_zen': 61.51355743408203, 'zcp_epe_nas': 7.4551044604106815, 'zcp_fisher': 0.1117168441414833, 'zcp_flops': 65121920.0, 'zcp_grad_norm': 21.642303466796875, 'zcp_grasp': -0.13195419311523438, 'zcp_jacov': -16.05928410527595, 'zcp_l2_norm': 560.9459228515625, 'zcp_nwot': 209.37091126893, 'zcp_params': 1936860.0, 'zcp_plain': 0.005565929692238569, 'zcp_snip': 37.80355453491211, 'lat_1080ti_1': 0.37246010316819117, 'lat_1080ti_32': 0.4507295920920529, 'lat_1080ti_64': 0.3698291341825981, 'lat_2080ti_1': 0.4408220350350519, 'lat_2080ti_32': 0.4144221062496081, 'lat_2080ti_64': 0.3718377804716541, 'lat_essential_ph_1': 0.1320754716981132, 'lat_eyeriss': 0.37197660132396904, 'lat_fpga': 0.44306791082436975, 'lat_gold_6226': 0.30553940823418974, 'lat_gold_6240': 0.4107926498980256, 'lat_pixel2': 0.2391304347826087, 'lat_pixel3': 0.4042157433616637, 'lat_raspi4': 0.44991896207634907, 'lat_samsung_a50': 0.14736842105263157, 'lat_samsung_s7': 0.47244094488188976, 'lat_silver_4114': 0.45425282006925516, 'lat_silver_4210r': 0.46022391658597156, 'lat_titan_rtx_1': 0.41379593197338815, 'lat_titan_rtx_32': 0.39941263704770363, 'lat_titan_rtx_64': 0.3859700962938957, 'lat_titanx_1': 0.22052831417476282, 'lat_titanx_32': 0.38659171831110933, 'lat_titanx_64': 0.34893262669233, 'lat_titanxp_1': 0.43162348429976755, 'lat_titanxp_32': 0.4151303041801844, 'lat_titanxp_64': 0.380698379294155}
| |
FBNet_2820
|
FBNet
|
2820
|
2820
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_696[FLOAT, 16x3x3x3]
%onnx::Conv_697[FLOAT, 16]
%onnx::Conv_699[FLOAT, 96x16x1x1]
%onnx::Conv_700[FLOAT, 96]
%onnx::Conv_702[FLOAT, 96x1x5x5]
%onnx::Conv_705[FLOAT, 16x96x1x1]
%onnx::Conv_708[FLOAT, 48x16x1x1]
%onnx::Conv_709[FLOAT, 48]
%onnx::Conv_711[FLOAT, 48x1x5x5]
%onnx::Conv_714[FLOAT, 24x48x1x1]
%onnx::Conv_715[FLOAT, 24]
%onnx::Conv_717[FLOAT, 24x12x1x1]
%onnx::Conv_720[FLOAT, 24x1x5x5]
%onnx::Conv_723[FLOAT, 24x12x1x1]
%onnx::Conv_726[FLOAT, 24x24x1x1]
%onnx::Conv_729[FLOAT, 24x1x5x5]
%onnx::Conv_732[FLOAT, 24x24x1x1]
%onnx::Conv_735[FLOAT, 144x24x1x1]
%onnx::Conv_736[FLOAT, 144]
%onnx::Conv_738[FLOAT, 144x1x5x5]
%onnx::Conv_741[FLOAT, 24x144x1x1]
%onnx::Conv_744[FLOAT, 24x24x1x1]
%onnx::Conv_747[FLOAT, 24x1x3x3]
%onnx::Conv_750[FLOAT, 32x24x1x1]
%onnx::Conv_751[FLOAT, 32]
%onnx::Conv_753[FLOAT, 32x16x1x1]
%onnx::Conv_756[FLOAT, 32x1x3x3]
%onnx::Conv_759[FLOAT, 32x16x1x1]
%onnx::Conv_762[FLOAT, 32x16x1x1]
%onnx::Conv_765[FLOAT, 32x1x3x3]
%onnx::Conv_768[FLOAT, 32x16x1x1]
%onnx::Conv_771[FLOAT, 96x32x1x1]
%onnx::Conv_774[FLOAT, 96x1x5x5]
%onnx::Conv_777[FLOAT, 32x96x1x1]
%onnx::Conv_780[FLOAT, 192x32x1x1]
%onnx::Conv_781[FLOAT, 192]
%onnx::Conv_783[FLOAT, 192x1x3x3]
%onnx::Conv_786[FLOAT, 64x192x1x1]
%onnx::Conv_787[FLOAT, 64]
%onnx::Conv_789[FLOAT, 64x64x1x1]
%onnx::Conv_792[FLOAT, 64x1x5x5]
%onnx::Conv_795[FLOAT, 64x64x1x1]
%onnx::Conv_798[FLOAT, 64x64x1x1]
%onnx::Conv_801[FLOAT, 64x1x5x5]
%onnx::Conv_804[FLOAT, 64x64x1x1]
%onnx::Conv_807[FLOAT, 64x32x1x1]
%onnx::Conv_810[FLOAT, 64x1x3x3]
%onnx::Conv_813[FLOAT, 112x32x1x1]
%onnx::Conv_814[FLOAT, 112]
%onnx::Conv_816[FLOAT, 112x56x1x1]
%onnx::Conv_819[FLOAT, 112x1x3x3]
%onnx::Conv_822[FLOAT, 112x56x1x1]
%onnx::Conv_825[FLOAT, 112x112x1x1]
%onnx::Conv_828[FLOAT, 112x1x5x5]
%onnx::Conv_831[FLOAT, 112x112x1x1]
%onnx::Conv_834[FLOAT, 336x112x1x1]
%onnx::Conv_835[FLOAT, 336]
%onnx::Conv_837[FLOAT, 336x1x5x5]
%onnx::Conv_840[FLOAT, 112x336x1x1]
%onnx::Conv_843[FLOAT, 336x112x1x1]
%onnx::Conv_846[FLOAT, 336x1x5x5]
%onnx::Conv_849[FLOAT, 184x336x1x1]
%onnx::Conv_850[FLOAT, 184]
%onnx::Conv_852[FLOAT, 184x92x1x1]
%onnx::Conv_855[FLOAT, 184x1x3x3]
%onnx::Conv_858[FLOAT, 184x92x1x1]
%onnx::Conv_861[FLOAT, 552x184x1x1]
%onnx::Conv_862[FLOAT, 552]
%onnx::Conv_864[FLOAT, 552x1x5x5]
%onnx::Conv_867[FLOAT, 184x552x1x1]
%onnx::Conv_870[FLOAT, 184x92x1x1]
%onnx::Conv_873[FLOAT, 184x1x3x3]
%onnx::Conv_876[FLOAT, 352x92x1x1]
%onnx::Conv_877[FLOAT, 352]
%onnx::Conv_879[FLOAT, 1504x352x1x1]
%onnx::Conv_880[FLOAT, 1504]
) {
%onnx::Conv_874 = Identity(%onnx::Conv_850)
%onnx::Conv_871 = Identity(%onnx::Conv_850)
%onnx::Conv_868 = Identity(%onnx::Conv_850)
%onnx::Conv_865 = Identity(%onnx::Conv_862)
%onnx::Conv_859 = Identity(%onnx::Conv_850)
%onnx::Conv_856 = Identity(%onnx::Conv_850)
%onnx::Conv_853 = Identity(%onnx::Conv_850)
%onnx::Conv_847 = Identity(%onnx::Conv_835)
%onnx::Conv_844 = Identity(%onnx::Conv_835)
%onnx::Conv_841 = Identity(%onnx::Conv_814)
%onnx::Conv_838 = Identity(%onnx::Conv_835)
%onnx::Conv_832 = Identity(%onnx::Conv_814)
%onnx::Conv_829 = Identity(%onnx::Conv_814)
%onnx::Conv_826 = Identity(%onnx::Conv_814)
%onnx::Conv_823 = Identity(%onnx::Conv_814)
%onnx::Conv_820 = Identity(%onnx::Conv_814)
%onnx::Conv_817 = Identity(%onnx::Conv_814)
%onnx::Conv_811 = Identity(%onnx::Conv_787)
%onnx::Conv_808 = Identity(%onnx::Conv_787)
%onnx::Conv_805 = Identity(%onnx::Conv_787)
%onnx::Conv_802 = Identity(%onnx::Conv_787)
%onnx::Conv_799 = Identity(%onnx::Conv_787)
%onnx::Conv_796 = Identity(%onnx::Conv_787)
%onnx::Conv_793 = Identity(%onnx::Conv_787)
%onnx::Conv_790 = Identity(%onnx::Conv_787)
%onnx::Conv_784 = Identity(%onnx::Conv_781)
%onnx::Conv_778 = Identity(%onnx::Conv_751)
%onnx::Conv_775 = Identity(%onnx::Conv_700)
%onnx::Conv_772 = Identity(%onnx::Conv_700)
%onnx::Conv_769 = Identity(%onnx::Conv_751)
%onnx::Conv_766 = Identity(%onnx::Conv_751)
%onnx::Conv_763 = Identity(%onnx::Conv_751)
%onnx::Conv_760 = Identity(%onnx::Conv_751)
%onnx::Conv_757 = Identity(%onnx::Conv_751)
%onnx::Conv_754 = Identity(%onnx::Conv_751)
%onnx::Conv_748 = Identity(%onnx::Conv_715)
%onnx::Conv_745 = Identity(%onnx::Conv_715)
%onnx::Conv_742 = Identity(%onnx::Conv_715)
%onnx::Conv_739 = Identity(%onnx::Conv_736)
%onnx::Conv_733 = Identity(%onnx::Conv_715)
%onnx::Conv_730 = Identity(%onnx::Conv_715)
%onnx::Conv_727 = Identity(%onnx::Conv_715)
%onnx::Conv_724 = Identity(%onnx::Conv_715)
%onnx::Conv_721 = Identity(%onnx::Conv_715)
%onnx::Conv_718 = Identity(%onnx::Conv_715)
%onnx::Conv_712 = Identity(%onnx::Conv_709)
%onnx::Conv_706 = Identity(%onnx::Conv_697)
%onnx::Conv_703 = Identity(%onnx::Conv_700)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_696, %onnx::Conv_697)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_output_0 = Reshape(%/cells.18/nl/Relu_output_0, %/cells.18/shuffle/Constant_output_0)
%/cells.18/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.18/shuffle/Reshape_output_0)
%/cells.18/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_1_output_0 = Reshape(%/cells.18/shuffle/Transpose_output_0, %/cells.18/shuffle/Constant_1_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/shuffle/Reshape_1_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_858, %onnx::Conv_859)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_861, %onnx::Conv_862)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_864, %onnx::Conv_865)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_867, %onnx::Conv_868)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_870, %onnx::Conv_871)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_873, %onnx::Conv_874)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_876, %onnx::Conv_877)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_879, %onnx::Conv_880)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%694 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %694
}
|
val_accuracy
| 0
| 58,891,136
| 1,314,916
|
{'zcp_synflow': 73.69486031604706, 'zcp_zen': 63.28453063964844, 'zcp_epe_nas': 15.828829114061485, 'zcp_fisher': 0.16123022139072418, 'zcp_flops': 58891136.0, 'zcp_grad_norm': 28.749162673950195, 'zcp_grasp': 0.06723785400390625, 'zcp_jacov': -16.05839084352566, 'zcp_l2_norm': 527.19189453125, 'zcp_nwot': 212.51792712633295, 'zcp_params': 1314916.0, 'zcp_plain': 0.0007517386693507433, 'zcp_snip': 41.83197021484375, 'lat_1080ti_1': 0.6312084996360791, 'lat_1080ti_32': 0.658328175059084, 'lat_1080ti_64': 0.5974383042560554, 'lat_2080ti_1': 0.6637741128462065, 'lat_2080ti_32': 0.6747710587939127, 'lat_2080ti_64': 0.5695989718490646, 'lat_essential_ph_1': 0.11320754716981132, 'lat_eyeriss': 0.36701630205563596, 'lat_fpga': 0.27914386645424943, 'lat_gold_6226': 0.18820710866985046, 'lat_gold_6240': 0.3800183998534342, 'lat_pixel2': 0.30434782608695654, 'lat_pixel3': 0.4456257840156126, 'lat_raspi4': 0.4306088473541048, 'lat_samsung_a50': 0.1368421052631579, 'lat_samsung_s7': 0.11023622047244094, 'lat_silver_4114': 0.5443460147537938, 'lat_silver_4210r': 0.4883339266119305, 'lat_titan_rtx_1': 0.6268335657968438, 'lat_titan_rtx_32': 0.6423894693282765, 'lat_titan_rtx_64': 0.6292575805406443, 'lat_titanx_1': 0.3238830088385389, 'lat_titanx_32': 0.6556629483772123, 'lat_titanx_64': 0.6205082735286588, 'lat_titanxp_1': 0.5775415148035123, 'lat_titanxp_32': 0.6711301642140808, 'lat_titanxp_64': 0.5930666682978699}
| |
FBNet_934
|
FBNet
|
934
|
934
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_760[FLOAT, 16x3x3x3]
%onnx::Conv_761[FLOAT, 16]
%onnx::Conv_763[FLOAT, 96x16x1x1]
%onnx::Conv_764[FLOAT, 96]
%onnx::Conv_766[FLOAT, 96x1x3x3]
%onnx::Conv_769[FLOAT, 16x96x1x1]
%onnx::Conv_772[FLOAT, 16x8x1x1]
%onnx::Conv_775[FLOAT, 16x1x3x3]
%onnx::Conv_778[FLOAT, 24x8x1x1]
%onnx::Conv_779[FLOAT, 24]
%onnx::Conv_781[FLOAT, 24x24x1x1]
%onnx::Conv_784[FLOAT, 24x1x5x5]
%onnx::Conv_787[FLOAT, 24x24x1x1]
%onnx::Conv_790[FLOAT, 24x12x1x1]
%onnx::Conv_793[FLOAT, 24x1x5x5]
%onnx::Conv_796[FLOAT, 24x12x1x1]
%onnx::Conv_799[FLOAT, 24x12x1x1]
%onnx::Conv_802[FLOAT, 24x1x5x5]
%onnx::Conv_805[FLOAT, 24x12x1x1]
%onnx::Conv_808[FLOAT, 24x24x1x1]
%onnx::Conv_811[FLOAT, 24x1x3x3]
%onnx::Conv_814[FLOAT, 32x24x1x1]
%onnx::Conv_815[FLOAT, 32]
%onnx::Conv_817[FLOAT, 96x32x1x1]
%onnx::Conv_820[FLOAT, 96x1x5x5]
%onnx::Conv_823[FLOAT, 32x96x1x1]
%onnx::Conv_826[FLOAT, 192x32x1x1]
%onnx::Conv_827[FLOAT, 192]
%onnx::Conv_829[FLOAT, 192x1x3x3]
%onnx::Conv_832[FLOAT, 32x192x1x1]
%onnx::Conv_835[FLOAT, 32x32x1x1]
%onnx::Conv_838[FLOAT, 32x1x5x5]
%onnx::Conv_841[FLOAT, 32x32x1x1]
%onnx::Conv_844[FLOAT, 32x16x1x1]
%onnx::Conv_847[FLOAT, 32x1x5x5]
%onnx::Conv_850[FLOAT, 64x16x1x1]
%onnx::Conv_851[FLOAT, 64]
%onnx::Conv_853[FLOAT, 64x64x1x1]
%onnx::Conv_856[FLOAT, 64x1x3x3]
%onnx::Conv_859[FLOAT, 64x64x1x1]
%onnx::Conv_862[FLOAT, 64x32x1x1]
%onnx::Conv_865[FLOAT, 64x1x3x3]
%onnx::Conv_868[FLOAT, 64x32x1x1]
%onnx::Conv_871[FLOAT, 64x64x1x1]
%onnx::Conv_874[FLOAT, 64x1x3x3]
%onnx::Conv_877[FLOAT, 112x64x1x1]
%onnx::Conv_878[FLOAT, 112]
%onnx::Conv_880[FLOAT, 112x56x1x1]
%onnx::Conv_883[FLOAT, 112x1x5x5]
%onnx::Conv_886[FLOAT, 112x56x1x1]
%onnx::Conv_889[FLOAT, 112x112x1x1]
%onnx::Conv_892[FLOAT, 112x1x5x5]
%onnx::Conv_895[FLOAT, 112x112x1x1]
%onnx::Conv_898[FLOAT, 112x56x1x1]
%onnx::Conv_901[FLOAT, 112x1x5x5]
%onnx::Conv_904[FLOAT, 112x56x1x1]
%onnx::Conv_907[FLOAT, 112x56x1x1]
%onnx::Conv_910[FLOAT, 112x1x3x3]
%onnx::Conv_913[FLOAT, 184x56x1x1]
%onnx::Conv_914[FLOAT, 184]
%onnx::Conv_916[FLOAT, 184x184x1x1]
%onnx::Conv_919[FLOAT, 184x1x3x3]
%onnx::Conv_922[FLOAT, 184x184x1x1]
%onnx::Conv_925[FLOAT, 184x92x1x1]
%onnx::Conv_928[FLOAT, 184x1x3x3]
%onnx::Conv_931[FLOAT, 184x92x1x1]
%onnx::Conv_934[FLOAT, 1104x184x1x1]
%onnx::Conv_935[FLOAT, 1104]
%onnx::Conv_937[FLOAT, 1104x1x3x3]
%onnx::Conv_940[FLOAT, 184x1104x1x1]
%onnx::Conv_943[FLOAT, 184x184x1x1]
%onnx::Conv_946[FLOAT, 184x1x5x5]
%onnx::Conv_949[FLOAT, 352x184x1x1]
%onnx::Conv_950[FLOAT, 352]
%onnx::Conv_952[FLOAT, 1504x352x1x1]
%onnx::Conv_953[FLOAT, 1504]
) {
%onnx::Conv_947 = Identity(%onnx::Conv_914)
%onnx::Conv_944 = Identity(%onnx::Conv_914)
%onnx::Conv_941 = Identity(%onnx::Conv_914)
%onnx::Conv_938 = Identity(%onnx::Conv_935)
%onnx::Conv_932 = Identity(%onnx::Conv_914)
%onnx::Conv_929 = Identity(%onnx::Conv_914)
%onnx::Conv_926 = Identity(%onnx::Conv_914)
%onnx::Conv_923 = Identity(%onnx::Conv_914)
%onnx::Conv_920 = Identity(%onnx::Conv_914)
%onnx::Conv_917 = Identity(%onnx::Conv_914)
%onnx::Conv_911 = Identity(%onnx::Conv_878)
%onnx::Conv_908 = Identity(%onnx::Conv_878)
%onnx::Conv_905 = Identity(%onnx::Conv_878)
%onnx::Conv_902 = Identity(%onnx::Conv_878)
%onnx::Conv_899 = Identity(%onnx::Conv_878)
%onnx::Conv_896 = Identity(%onnx::Conv_878)
%onnx::Conv_893 = Identity(%onnx::Conv_878)
%onnx::Conv_890 = Identity(%onnx::Conv_878)
%onnx::Conv_887 = Identity(%onnx::Conv_878)
%onnx::Conv_884 = Identity(%onnx::Conv_878)
%onnx::Conv_881 = Identity(%onnx::Conv_878)
%onnx::Conv_875 = Identity(%onnx::Conv_851)
%onnx::Conv_872 = Identity(%onnx::Conv_851)
%onnx::Conv_869 = Identity(%onnx::Conv_851)
%onnx::Conv_866 = Identity(%onnx::Conv_851)
%onnx::Conv_863 = Identity(%onnx::Conv_851)
%onnx::Conv_860 = Identity(%onnx::Conv_851)
%onnx::Conv_857 = Identity(%onnx::Conv_851)
%onnx::Conv_854 = Identity(%onnx::Conv_851)
%onnx::Conv_848 = Identity(%onnx::Conv_815)
%onnx::Conv_845 = Identity(%onnx::Conv_815)
%onnx::Conv_842 = Identity(%onnx::Conv_815)
%onnx::Conv_839 = Identity(%onnx::Conv_815)
%onnx::Conv_836 = Identity(%onnx::Conv_815)
%onnx::Conv_833 = Identity(%onnx::Conv_815)
%onnx::Conv_830 = Identity(%onnx::Conv_827)
%onnx::Conv_824 = Identity(%onnx::Conv_815)
%onnx::Conv_821 = Identity(%onnx::Conv_764)
%onnx::Conv_818 = Identity(%onnx::Conv_764)
%onnx::Conv_812 = Identity(%onnx::Conv_779)
%onnx::Conv_809 = Identity(%onnx::Conv_779)
%onnx::Conv_806 = Identity(%onnx::Conv_779)
%onnx::Conv_803 = Identity(%onnx::Conv_779)
%onnx::Conv_800 = Identity(%onnx::Conv_779)
%onnx::Conv_797 = Identity(%onnx::Conv_779)
%onnx::Conv_794 = Identity(%onnx::Conv_779)
%onnx::Conv_791 = Identity(%onnx::Conv_779)
%onnx::Conv_788 = Identity(%onnx::Conv_779)
%onnx::Conv_785 = Identity(%onnx::Conv_779)
%onnx::Conv_782 = Identity(%onnx::Conv_779)
%onnx::Conv_776 = Identity(%onnx::Conv_761)
%onnx::Conv_773 = Identity(%onnx::Conv_761)
%onnx::Conv_770 = Identity(%onnx::Conv_761)
%onnx::Conv_767 = Identity(%onnx::Conv_764)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_760, %onnx::Conv_761)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_823, %onnx::Conv_824)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_826, %onnx::Conv_827)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_829, %onnx::Conv_830)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_832, %onnx::Conv_833)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_835, %onnx::Conv_836)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_838, %onnx::Conv_839)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_841, %onnx::Conv_842)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_844, %onnx::Conv_845)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_847, %onnx::Conv_848)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_850, %onnx::Conv_851)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_853, %onnx::Conv_854)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_856, %onnx::Conv_857)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_859, %onnx::Conv_860)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_862, %onnx::Conv_863)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_865, %onnx::Conv_866)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_868, %onnx::Conv_869)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_871, %onnx::Conv_872)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_874, %onnx::Conv_875)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_877, %onnx::Conv_878)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_880, %onnx::Conv_881)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_883, %onnx::Conv_884)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_886, %onnx::Conv_887)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_889, %onnx::Conv_890)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_892, %onnx::Conv_893)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_895, %onnx::Conv_896)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_898, %onnx::Conv_899)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_901, %onnx::Conv_902)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_904, %onnx::Conv_905)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_907, %onnx::Conv_908)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_910, %onnx::Conv_911)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_913, %onnx::Conv_914)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_916, %onnx::Conv_917)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_919, %onnx::Conv_920)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_922, %onnx::Conv_923)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_925, %onnx::Conv_926)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_928, %onnx::Conv_929)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_931, %onnx::Conv_932)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_934, %onnx::Conv_935)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_937, %onnx::Conv_938)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_940, %onnx::Conv_941)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_943, %onnx::Conv_944)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_946, %onnx::Conv_947)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_949, %onnx::Conv_950)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_952, %onnx::Conv_953)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%758 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %758
}
|
val_accuracy
| 0
| 43,623,552
| 1,462,300
|
{'zcp_synflow': 73.74577930645066, 'zcp_zen': 64.08456420898438, 'zcp_epe_nas': 12.052201937624515, 'zcp_fisher': 0.17677146196365356, 'zcp_flops': 43623552.0, 'zcp_grad_norm': 24.875125885009766, 'zcp_grasp': -0.032196044921875, 'zcp_jacov': -16.062129185063405, 'zcp_l2_norm': 534.9183959960938, 'zcp_nwot': 206.98750268849187, 'zcp_params': 1462300.0, 'zcp_plain': -0.008535937406122684, 'zcp_snip': 43.793949127197266, 'lat_1080ti_1': 0.7035360926720224, 'lat_1080ti_32': 0.6254084414345998, 'lat_1080ti_64': 0.3722497806372126, 'lat_2080ti_1': 0.7861691800676981, 'lat_2080ti_32': 0.6513772042541387, 'lat_2080ti_64': 0.4172015600427908, 'lat_essential_ph_1': 0.2830188679245283, 'lat_eyeriss': 0.20901105752480145, 'lat_fpga': 0.17206780197706287, 'lat_gold_6226': 0.11490874236373425, 'lat_gold_6240': 0.45355680248174746, 'lat_pixel2': 0.2391304347826087, 'lat_pixel3': 0.21708109635015194, 'lat_raspi4': 0.2673642014118921, 'lat_samsung_a50': 0.10526315789473684, 'lat_samsung_s7': 0.14960629921259844, 'lat_silver_4114': 0.5443510169385152, 'lat_silver_4210r': 0.5458485705336004, 'lat_titan_rtx_1': 0.7554930120318445, 'lat_titan_rtx_32': 0.6533623810327035, 'lat_titan_rtx_64': 0.4766283098928778, 'lat_titanx_1': 0.40559914595330465, 'lat_titanx_32': 0.5483256252758761, 'lat_titanx_64': 0.36313716704195736, 'lat_titanxp_1': 0.697960309213921, 'lat_titanxp_32': 0.5988978262075715, 'lat_titanxp_64': 0.41629828594781343}
| |
FBNet_1999
|
FBNet
|
1999
|
1999
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_678[FLOAT, 16x3x3x3]
%onnx::Conv_679[FLOAT, 16]
%onnx::Conv_681[FLOAT, 16x8x1x1]
%onnx::Conv_684[FLOAT, 16x1x3x3]
%onnx::Conv_687[FLOAT, 16x8x1x1]
%onnx::Conv_690[FLOAT, 48x16x1x1]
%onnx::Conv_691[FLOAT, 48]
%onnx::Conv_693[FLOAT, 48x1x5x5]
%onnx::Conv_696[FLOAT, 24x48x1x1]
%onnx::Conv_697[FLOAT, 24]
%onnx::Conv_699[FLOAT, 24x12x1x1]
%onnx::Conv_702[FLOAT, 24x1x5x5]
%onnx::Conv_705[FLOAT, 24x12x1x1]
%onnx::Conv_708[FLOAT, 144x24x1x1]
%onnx::Conv_709[FLOAT, 144]
%onnx::Conv_711[FLOAT, 144x1x3x3]
%onnx::Conv_714[FLOAT, 24x144x1x1]
%onnx::Conv_717[FLOAT, 72x24x1x1]
%onnx::Conv_718[FLOAT, 72]
%onnx::Conv_720[FLOAT, 72x1x3x3]
%onnx::Conv_723[FLOAT, 32x72x1x1]
%onnx::Conv_724[FLOAT, 32]
%onnx::Conv_726[FLOAT, 32x16x1x1]
%onnx::Conv_729[FLOAT, 32x1x5x5]
%onnx::Conv_732[FLOAT, 32x16x1x1]
%onnx::Conv_735[FLOAT, 96x32x1x1]
%onnx::Conv_736[FLOAT, 96]
%onnx::Conv_738[FLOAT, 96x1x3x3]
%onnx::Conv_741[FLOAT, 32x96x1x1]
%onnx::Conv_744[FLOAT, 96x32x1x1]
%onnx::Conv_747[FLOAT, 96x1x3x3]
%onnx::Conv_750[FLOAT, 32x96x1x1]
%onnx::Conv_753[FLOAT, 32x16x1x1]
%onnx::Conv_756[FLOAT, 32x1x3x3]
%onnx::Conv_759[FLOAT, 64x16x1x1]
%onnx::Conv_760[FLOAT, 64]
%onnx::Conv_762[FLOAT, 64x32x1x1]
%onnx::Conv_765[FLOAT, 64x1x5x5]
%onnx::Conv_768[FLOAT, 64x32x1x1]
%onnx::Conv_771[FLOAT, 64x64x1x1]
%onnx::Conv_774[FLOAT, 64x1x3x3]
%onnx::Conv_777[FLOAT, 64x64x1x1]
%onnx::Conv_780[FLOAT, 384x64x1x1]
%onnx::Conv_781[FLOAT, 384]
%onnx::Conv_783[FLOAT, 384x1x5x5]
%onnx::Conv_786[FLOAT, 64x384x1x1]
%onnx::Conv_789[FLOAT, 64x64x1x1]
%onnx::Conv_792[FLOAT, 64x1x5x5]
%onnx::Conv_795[FLOAT, 112x64x1x1]
%onnx::Conv_796[FLOAT, 112]
%onnx::Conv_798[FLOAT, 112x56x1x1]
%onnx::Conv_801[FLOAT, 112x1x5x5]
%onnx::Conv_804[FLOAT, 112x56x1x1]
%onnx::Conv_807[FLOAT, 336x112x1x1]
%onnx::Conv_808[FLOAT, 336]
%onnx::Conv_810[FLOAT, 336x1x5x5]
%onnx::Conv_813[FLOAT, 112x336x1x1]
%onnx::Conv_816[FLOAT, 112x112x1x1]
%onnx::Conv_819[FLOAT, 112x1x3x3]
%onnx::Conv_822[FLOAT, 112x112x1x1]
%onnx::Conv_825[FLOAT, 672x112x1x1]
%onnx::Conv_826[FLOAT, 672]
%onnx::Conv_828[FLOAT, 672x1x5x5]
%onnx::Conv_831[FLOAT, 184x672x1x1]
%onnx::Conv_832[FLOAT, 184]
%onnx::Conv_834[FLOAT, 1104x184x1x1]
%onnx::Conv_835[FLOAT, 1104]
%onnx::Conv_837[FLOAT, 1104x1x5x5]
%onnx::Conv_840[FLOAT, 184x1104x1x1]
%onnx::Conv_843[FLOAT, 1104x184x1x1]
%onnx::Conv_846[FLOAT, 1104x1x5x5]
%onnx::Conv_849[FLOAT, 184x1104x1x1]
%onnx::Conv_852[FLOAT, 1104x184x1x1]
%onnx::Conv_855[FLOAT, 1104x1x3x3]
%onnx::Conv_858[FLOAT, 352x1104x1x1]
%onnx::Conv_859[FLOAT, 352]
%onnx::Conv_861[FLOAT, 1504x352x1x1]
%onnx::Conv_862[FLOAT, 1504]
) {
%onnx::Conv_856 = Identity(%onnx::Conv_835)
%onnx::Conv_853 = Identity(%onnx::Conv_835)
%onnx::Conv_850 = Identity(%onnx::Conv_832)
%onnx::Conv_847 = Identity(%onnx::Conv_835)
%onnx::Conv_844 = Identity(%onnx::Conv_835)
%onnx::Conv_841 = Identity(%onnx::Conv_832)
%onnx::Conv_838 = Identity(%onnx::Conv_835)
%onnx::Conv_829 = Identity(%onnx::Conv_826)
%onnx::Conv_823 = Identity(%onnx::Conv_796)
%onnx::Conv_820 = Identity(%onnx::Conv_796)
%onnx::Conv_817 = Identity(%onnx::Conv_796)
%onnx::Conv_814 = Identity(%onnx::Conv_796)
%onnx::Conv_811 = Identity(%onnx::Conv_808)
%onnx::Conv_805 = Identity(%onnx::Conv_796)
%onnx::Conv_802 = Identity(%onnx::Conv_796)
%onnx::Conv_799 = Identity(%onnx::Conv_796)
%onnx::Conv_793 = Identity(%onnx::Conv_760)
%onnx::Conv_790 = Identity(%onnx::Conv_760)
%onnx::Conv_787 = Identity(%onnx::Conv_760)
%onnx::Conv_784 = Identity(%onnx::Conv_781)
%onnx::Conv_778 = Identity(%onnx::Conv_760)
%onnx::Conv_775 = Identity(%onnx::Conv_760)
%onnx::Conv_772 = Identity(%onnx::Conv_760)
%onnx::Conv_769 = Identity(%onnx::Conv_760)
%onnx::Conv_766 = Identity(%onnx::Conv_760)
%onnx::Conv_763 = Identity(%onnx::Conv_760)
%onnx::Conv_757 = Identity(%onnx::Conv_724)
%onnx::Conv_754 = Identity(%onnx::Conv_724)
%onnx::Conv_751 = Identity(%onnx::Conv_724)
%onnx::Conv_748 = Identity(%onnx::Conv_736)
%onnx::Conv_745 = Identity(%onnx::Conv_736)
%onnx::Conv_742 = Identity(%onnx::Conv_724)
%onnx::Conv_739 = Identity(%onnx::Conv_736)
%onnx::Conv_733 = Identity(%onnx::Conv_724)
%onnx::Conv_730 = Identity(%onnx::Conv_724)
%onnx::Conv_727 = Identity(%onnx::Conv_724)
%onnx::Conv_721 = Identity(%onnx::Conv_718)
%onnx::Conv_715 = Identity(%onnx::Conv_697)
%onnx::Conv_712 = Identity(%onnx::Conv_709)
%onnx::Conv_706 = Identity(%onnx::Conv_697)
%onnx::Conv_703 = Identity(%onnx::Conv_697)
%onnx::Conv_700 = Identity(%onnx::Conv_697)
%onnx::Conv_694 = Identity(%onnx::Conv_691)
%onnx::Conv_688 = Identity(%onnx::Conv_679)
%onnx::Conv_685 = Identity(%onnx::Conv_679)
%onnx::Conv_682 = Identity(%onnx::Conv_679)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_678, %onnx::Conv_679)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_858, %onnx::Conv_859)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_861, %onnx::Conv_862)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%676 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %676
}
|
val_accuracy
| 0
| 76,119,680
| 2,641,316
|
{'zcp_synflow': 74.4168751042752, 'zcp_zen': 67.1103515625, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.07579173147678375, 'zcp_flops': 76119680.0, 'zcp_grad_norm': 20.9899845123291, 'zcp_grasp': -0.09354019165039062, 'zcp_jacov': -16.054607549925727, 'zcp_l2_norm': 642.0810546875, 'zcp_nwot': 212.0959543487143, 'zcp_params': 2641316.0, 'zcp_plain': -0.003522830782458186, 'zcp_snip': 39.790767669677734, 'lat_1080ti_1': 0.6202648468492076, 'lat_1080ti_32': 0.5338043238902532, 'lat_1080ti_64': 0.4074183194945563, 'lat_2080ti_1': 0.6383608406575709, 'lat_2080ti_32': 0.5623614545499888, 'lat_2080ti_64': 0.44211981484754664, 'lat_essential_ph_1': 0.5283018867924528, 'lat_eyeriss': 0.535675645939156, 'lat_fpga': 0.5666689755489367, 'lat_gold_6226': 0.5072432148474246, 'lat_gold_6240': 0.6457406834081028, 'lat_pixel2': 0.5217391304347826, 'lat_pixel3': 0.5156554431378917, 'lat_raspi4': 0.6183681178329549, 'lat_samsung_a50': 0.3473684210526316, 'lat_samsung_s7': 0.29133858267716534, 'lat_silver_4114': 0.6356061075753447, 'lat_silver_4210r': 0.6742849484066564, 'lat_titan_rtx_1': 0.6057009229204218, 'lat_titan_rtx_32': 0.5504829737886656, 'lat_titan_rtx_64': 0.46229854342550536, 'lat_titanx_1': 0.34063446767721073, 'lat_titanx_32': 0.5232087372615702, 'lat_titanx_64': 0.4054206047754837, 'lat_titanxp_1': 0.5920532727934222, 'lat_titanxp_32': 0.5478394526969221, 'lat_titanxp_64': 0.4237198469139218}
| |
FBNet_4411
|
FBNet
|
4411
|
4411
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_651[FLOAT, 16x3x3x3]
%onnx::Conv_652[FLOAT, 16]
%onnx::Conv_654[FLOAT, 16x16x1x1]
%onnx::Conv_657[FLOAT, 16x1x5x5]
%onnx::Conv_660[FLOAT, 16x16x1x1]
%onnx::Conv_663[FLOAT, 16x8x1x1]
%onnx::Conv_666[FLOAT, 16x1x5x5]
%onnx::Conv_669[FLOAT, 24x8x1x1]
%onnx::Conv_670[FLOAT, 24]
%onnx::Conv_672[FLOAT, 72x24x1x1]
%onnx::Conv_673[FLOAT, 72]
%onnx::Conv_675[FLOAT, 72x1x5x5]
%onnx::Conv_678[FLOAT, 24x72x1x1]
%onnx::Conv_681[FLOAT, 144x24x1x1]
%onnx::Conv_682[FLOAT, 144]
%onnx::Conv_684[FLOAT, 144x1x5x5]
%onnx::Conv_687[FLOAT, 24x144x1x1]
%onnx::Conv_690[FLOAT, 144x24x1x1]
%onnx::Conv_693[FLOAT, 144x1x3x3]
%onnx::Conv_696[FLOAT, 24x144x1x1]
%onnx::Conv_699[FLOAT, 24x24x1x1]
%onnx::Conv_702[FLOAT, 24x1x5x5]
%onnx::Conv_705[FLOAT, 32x24x1x1]
%onnx::Conv_706[FLOAT, 32]
%onnx::Conv_708[FLOAT, 32x32x1x1]
%onnx::Conv_711[FLOAT, 32x1x5x5]
%onnx::Conv_714[FLOAT, 32x32x1x1]
%onnx::Conv_717[FLOAT, 32x16x1x1]
%onnx::Conv_720[FLOAT, 32x1x5x5]
%onnx::Conv_723[FLOAT, 32x16x1x1]
%onnx::Conv_726[FLOAT, 192x32x1x1]
%onnx::Conv_727[FLOAT, 192]
%onnx::Conv_729[FLOAT, 192x1x5x5]
%onnx::Conv_732[FLOAT, 64x192x1x1]
%onnx::Conv_733[FLOAT, 64]
%onnx::Conv_735[FLOAT, 64x32x1x1]
%onnx::Conv_738[FLOAT, 64x1x3x3]
%onnx::Conv_741[FLOAT, 64x32x1x1]
%onnx::Conv_744[FLOAT, 384x64x1x1]
%onnx::Conv_745[FLOAT, 384]
%onnx::Conv_747[FLOAT, 384x1x3x3]
%onnx::Conv_750[FLOAT, 64x384x1x1]
%onnx::Conv_753[FLOAT, 384x64x1x1]
%onnx::Conv_756[FLOAT, 384x1x3x3]
%onnx::Conv_759[FLOAT, 64x384x1x1]
%onnx::Conv_762[FLOAT, 112x64x1x1]
%onnx::Conv_763[FLOAT, 112]
%onnx::Conv_765[FLOAT, 112x112x1x1]
%onnx::Conv_768[FLOAT, 112x1x5x5]
%onnx::Conv_771[FLOAT, 112x112x1x1]
%onnx::Conv_774[FLOAT, 672x112x1x1]
%onnx::Conv_775[FLOAT, 672]
%onnx::Conv_777[FLOAT, 672x1x5x5]
%onnx::Conv_780[FLOAT, 112x672x1x1]
%onnx::Conv_783[FLOAT, 672x112x1x1]
%onnx::Conv_786[FLOAT, 672x1x5x5]
%onnx::Conv_789[FLOAT, 112x672x1x1]
%onnx::Conv_792[FLOAT, 112x56x1x1]
%onnx::Conv_795[FLOAT, 112x1x3x3]
%onnx::Conv_798[FLOAT, 184x56x1x1]
%onnx::Conv_799[FLOAT, 184]
%onnx::Conv_801[FLOAT, 1104x184x1x1]
%onnx::Conv_802[FLOAT, 1104]
%onnx::Conv_804[FLOAT, 1104x1x5x5]
%onnx::Conv_807[FLOAT, 184x1104x1x1]
%onnx::Conv_810[FLOAT, 552x184x1x1]
%onnx::Conv_811[FLOAT, 552]
%onnx::Conv_813[FLOAT, 552x1x3x3]
%onnx::Conv_816[FLOAT, 184x552x1x1]
%onnx::Conv_819[FLOAT, 552x184x1x1]
%onnx::Conv_822[FLOAT, 552x1x3x3]
%onnx::Conv_825[FLOAT, 184x552x1x1]
%onnx::Conv_828[FLOAT, 552x184x1x1]
%onnx::Conv_831[FLOAT, 552x1x5x5]
%onnx::Conv_834[FLOAT, 352x552x1x1]
%onnx::Conv_835[FLOAT, 352]
%onnx::Conv_837[FLOAT, 1504x352x1x1]
%onnx::Conv_838[FLOAT, 1504]
) {
%onnx::Conv_832 = Identity(%onnx::Conv_811)
%onnx::Conv_829 = Identity(%onnx::Conv_811)
%onnx::Conv_826 = Identity(%onnx::Conv_799)
%onnx::Conv_823 = Identity(%onnx::Conv_811)
%onnx::Conv_820 = Identity(%onnx::Conv_811)
%onnx::Conv_817 = Identity(%onnx::Conv_799)
%onnx::Conv_814 = Identity(%onnx::Conv_811)
%onnx::Conv_808 = Identity(%onnx::Conv_799)
%onnx::Conv_805 = Identity(%onnx::Conv_802)
%onnx::Conv_796 = Identity(%onnx::Conv_763)
%onnx::Conv_793 = Identity(%onnx::Conv_763)
%onnx::Conv_790 = Identity(%onnx::Conv_763)
%onnx::Conv_787 = Identity(%onnx::Conv_775)
%onnx::Conv_784 = Identity(%onnx::Conv_775)
%onnx::Conv_781 = Identity(%onnx::Conv_763)
%onnx::Conv_778 = Identity(%onnx::Conv_775)
%onnx::Conv_772 = Identity(%onnx::Conv_763)
%onnx::Conv_769 = Identity(%onnx::Conv_763)
%onnx::Conv_766 = Identity(%onnx::Conv_763)
%onnx::Conv_760 = Identity(%onnx::Conv_733)
%onnx::Conv_757 = Identity(%onnx::Conv_745)
%onnx::Conv_754 = Identity(%onnx::Conv_745)
%onnx::Conv_751 = Identity(%onnx::Conv_733)
%onnx::Conv_748 = Identity(%onnx::Conv_745)
%onnx::Conv_742 = Identity(%onnx::Conv_733)
%onnx::Conv_739 = Identity(%onnx::Conv_733)
%onnx::Conv_736 = Identity(%onnx::Conv_733)
%onnx::Conv_730 = Identity(%onnx::Conv_727)
%onnx::Conv_724 = Identity(%onnx::Conv_706)
%onnx::Conv_721 = Identity(%onnx::Conv_706)
%onnx::Conv_718 = Identity(%onnx::Conv_706)
%onnx::Conv_715 = Identity(%onnx::Conv_706)
%onnx::Conv_712 = Identity(%onnx::Conv_706)
%onnx::Conv_709 = Identity(%onnx::Conv_706)
%onnx::Conv_703 = Identity(%onnx::Conv_670)
%onnx::Conv_700 = Identity(%onnx::Conv_670)
%onnx::Conv_697 = Identity(%onnx::Conv_670)
%onnx::Conv_694 = Identity(%onnx::Conv_682)
%onnx::Conv_691 = Identity(%onnx::Conv_682)
%onnx::Conv_688 = Identity(%onnx::Conv_670)
%onnx::Conv_685 = Identity(%onnx::Conv_682)
%onnx::Conv_679 = Identity(%onnx::Conv_670)
%onnx::Conv_676 = Identity(%onnx::Conv_673)
%onnx::Conv_667 = Identity(%onnx::Conv_652)
%onnx::Conv_664 = Identity(%onnx::Conv_652)
%onnx::Conv_661 = Identity(%onnx::Conv_652)
%onnx::Conv_658 = Identity(%onnx::Conv_652)
%onnx::Conv_655 = Identity(%onnx::Conv_652)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_651, %onnx::Conv_652)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.13/relu/Relu_output_0 = Relu(%/cells.13/conv/Conv_output_0)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/relu/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/relu/Relu_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%649 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %649
}
|
val_accuracy
| 0
| 93,741,696
| 2,423,100
|
{'zcp_synflow': 80.74619837629155, 'zcp_zen': 72.16517639160156, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.14400652050971985, 'zcp_flops': 93741696.0, 'zcp_grad_norm': 24.957361221313477, 'zcp_grasp': -0.05328178405761719, 'zcp_jacov': -16.052295472265286, 'zcp_l2_norm': 696.0731201171875, 'zcp_nwot': 216.63855890947698, 'zcp_params': 2423100.0, 'zcp_plain': 0.009139979258179665, 'zcp_snip': 49.2861213684082, 'lat_1080ti_1': 0.6695486580405972, 'lat_1080ti_32': 0.732349703359114, 'lat_1080ti_64': 0.6862171227439668, 'lat_2080ti_1': 0.6512212427189983, 'lat_2080ti_32': 0.7073846525518566, 'lat_2080ti_64': 0.6886734678618994, 'lat_essential_ph_1': 0.32075471698113206, 'lat_eyeriss': 0.7295215740927508, 'lat_fpga': 0.790162182487458, 'lat_gold_6226': 0.5781099439921824, 'lat_gold_6240': 0.6839292862764218, 'lat_pixel2': 0.45652173913043476, 'lat_pixel3': 0.7580104582802888, 'lat_raspi4': 0.8094102987513674, 'lat_samsung_a50': 0.3368421052631579, 'lat_samsung_s7': 0.1968503937007874, 'lat_silver_4114': 0.684862410257414, 'lat_silver_4210r': 0.727208152187635, 'lat_titan_rtx_1': 0.6156339178433207, 'lat_titan_rtx_32': 0.66654687421912, 'lat_titan_rtx_64': 0.7012311464266888, 'lat_titanx_1': 0.3331438734956259, 'lat_titanx_32': 0.7202728214918676, 'lat_titanx_64': 0.7072919481049397, 'lat_titanxp_1': 0.5800616630220491, 'lat_titanxp_32': 0.7173818039124067, 'lat_titanxp_64': 0.7031949348109753}
| |
FBNet_2985
|
FBNet
|
2985
|
2985
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_706[FLOAT, 16x3x3x3]
%onnx::Conv_707[FLOAT, 16]
%onnx::Conv_709[FLOAT, 48x16x1x1]
%onnx::Conv_710[FLOAT, 48]
%onnx::Conv_712[FLOAT, 48x1x5x5]
%onnx::Conv_715[FLOAT, 24x48x1x1]
%onnx::Conv_716[FLOAT, 24]
%onnx::Conv_718[FLOAT, 144x24x1x1]
%onnx::Conv_719[FLOAT, 144]
%onnx::Conv_721[FLOAT, 144x1x5x5]
%onnx::Conv_724[FLOAT, 24x144x1x1]
%onnx::Conv_727[FLOAT, 144x24x1x1]
%onnx::Conv_730[FLOAT, 144x1x3x3]
%onnx::Conv_733[FLOAT, 24x144x1x1]
%onnx::Conv_736[FLOAT, 24x12x1x1]
%onnx::Conv_739[FLOAT, 24x1x5x5]
%onnx::Conv_742[FLOAT, 24x12x1x1]
%onnx::Conv_745[FLOAT, 32x24x1x1]
%onnx::Conv_746[FLOAT, 32]
%onnx::Conv_748[FLOAT, 32x16x1x1]
%onnx::Conv_751[FLOAT, 32x1x5x5]
%onnx::Conv_754[FLOAT, 32x16x1x1]
%onnx::Conv_757[FLOAT, 192x32x1x1]
%onnx::Conv_758[FLOAT, 192]
%onnx::Conv_760[FLOAT, 192x1x3x3]
%onnx::Conv_763[FLOAT, 32x192x1x1]
%onnx::Conv_766[FLOAT, 96x32x1x1]
%onnx::Conv_767[FLOAT, 96]
%onnx::Conv_769[FLOAT, 96x1x5x5]
%onnx::Conv_772[FLOAT, 32x96x1x1]
%onnx::Conv_775[FLOAT, 192x32x1x1]
%onnx::Conv_778[FLOAT, 192x1x3x3]
%onnx::Conv_781[FLOAT, 64x192x1x1]
%onnx::Conv_782[FLOAT, 64]
%onnx::Conv_784[FLOAT, 192x64x1x1]
%onnx::Conv_787[FLOAT, 192x1x3x3]
%onnx::Conv_790[FLOAT, 64x192x1x1]
%onnx::Conv_793[FLOAT, 64x64x1x1]
%onnx::Conv_796[FLOAT, 64x1x3x3]
%onnx::Conv_799[FLOAT, 64x64x1x1]
%onnx::Conv_802[FLOAT, 192x64x1x1]
%onnx::Conv_805[FLOAT, 192x1x5x5]
%onnx::Conv_808[FLOAT, 64x192x1x1]
%onnx::Conv_811[FLOAT, 64x32x1x1]
%onnx::Conv_814[FLOAT, 64x1x5x5]
%onnx::Conv_817[FLOAT, 112x32x1x1]
%onnx::Conv_818[FLOAT, 112]
%onnx::Conv_820[FLOAT, 112x56x1x1]
%onnx::Conv_823[FLOAT, 112x1x5x5]
%onnx::Conv_826[FLOAT, 112x56x1x1]
%onnx::Conv_829[FLOAT, 112x112x1x1]
%onnx::Conv_832[FLOAT, 112x1x5x5]
%onnx::Conv_835[FLOAT, 112x112x1x1]
%onnx::Conv_838[FLOAT, 336x112x1x1]
%onnx::Conv_839[FLOAT, 336]
%onnx::Conv_841[FLOAT, 336x1x3x3]
%onnx::Conv_844[FLOAT, 112x336x1x1]
%onnx::Conv_847[FLOAT, 672x112x1x1]
%onnx::Conv_848[FLOAT, 672]
%onnx::Conv_850[FLOAT, 672x1x5x5]
%onnx::Conv_853[FLOAT, 184x672x1x1]
%onnx::Conv_854[FLOAT, 184]
%onnx::Conv_856[FLOAT, 184x92x1x1]
%onnx::Conv_859[FLOAT, 184x1x3x3]
%onnx::Conv_862[FLOAT, 184x92x1x1]
%onnx::Conv_865[FLOAT, 1104x184x1x1]
%onnx::Conv_866[FLOAT, 1104]
%onnx::Conv_868[FLOAT, 1104x1x5x5]
%onnx::Conv_871[FLOAT, 184x1104x1x1]
%onnx::Conv_874[FLOAT, 184x92x1x1]
%onnx::Conv_877[FLOAT, 184x1x3x3]
%onnx::Conv_880[FLOAT, 184x92x1x1]
%onnx::Conv_883[FLOAT, 184x92x1x1]
%onnx::Conv_886[FLOAT, 184x1x3x3]
%onnx::Conv_889[FLOAT, 352x92x1x1]
%onnx::Conv_890[FLOAT, 352]
%onnx::Conv_892[FLOAT, 1504x352x1x1]
%onnx::Conv_893[FLOAT, 1504]
) {
%onnx::Conv_887 = Identity(%onnx::Conv_854)
%onnx::Conv_884 = Identity(%onnx::Conv_854)
%onnx::Conv_881 = Identity(%onnx::Conv_854)
%onnx::Conv_878 = Identity(%onnx::Conv_854)
%onnx::Conv_875 = Identity(%onnx::Conv_854)
%onnx::Conv_872 = Identity(%onnx::Conv_854)
%onnx::Conv_869 = Identity(%onnx::Conv_866)
%onnx::Conv_863 = Identity(%onnx::Conv_854)
%onnx::Conv_860 = Identity(%onnx::Conv_854)
%onnx::Conv_857 = Identity(%onnx::Conv_854)
%onnx::Conv_851 = Identity(%onnx::Conv_848)
%onnx::Conv_845 = Identity(%onnx::Conv_818)
%onnx::Conv_842 = Identity(%onnx::Conv_839)
%onnx::Conv_836 = Identity(%onnx::Conv_818)
%onnx::Conv_833 = Identity(%onnx::Conv_818)
%onnx::Conv_830 = Identity(%onnx::Conv_818)
%onnx::Conv_827 = Identity(%onnx::Conv_818)
%onnx::Conv_824 = Identity(%onnx::Conv_818)
%onnx::Conv_821 = Identity(%onnx::Conv_818)
%onnx::Conv_815 = Identity(%onnx::Conv_782)
%onnx::Conv_812 = Identity(%onnx::Conv_782)
%onnx::Conv_809 = Identity(%onnx::Conv_782)
%onnx::Conv_806 = Identity(%onnx::Conv_758)
%onnx::Conv_803 = Identity(%onnx::Conv_758)
%onnx::Conv_800 = Identity(%onnx::Conv_782)
%onnx::Conv_797 = Identity(%onnx::Conv_782)
%onnx::Conv_794 = Identity(%onnx::Conv_782)
%onnx::Conv_791 = Identity(%onnx::Conv_782)
%onnx::Conv_788 = Identity(%onnx::Conv_758)
%onnx::Conv_785 = Identity(%onnx::Conv_758)
%onnx::Conv_779 = Identity(%onnx::Conv_758)
%onnx::Conv_776 = Identity(%onnx::Conv_758)
%onnx::Conv_773 = Identity(%onnx::Conv_746)
%onnx::Conv_770 = Identity(%onnx::Conv_767)
%onnx::Conv_764 = Identity(%onnx::Conv_746)
%onnx::Conv_761 = Identity(%onnx::Conv_758)
%onnx::Conv_755 = Identity(%onnx::Conv_746)
%onnx::Conv_752 = Identity(%onnx::Conv_746)
%onnx::Conv_749 = Identity(%onnx::Conv_746)
%onnx::Conv_743 = Identity(%onnx::Conv_716)
%onnx::Conv_740 = Identity(%onnx::Conv_716)
%onnx::Conv_737 = Identity(%onnx::Conv_716)
%onnx::Conv_734 = Identity(%onnx::Conv_716)
%onnx::Conv_731 = Identity(%onnx::Conv_719)
%onnx::Conv_728 = Identity(%onnx::Conv_719)
%onnx::Conv_725 = Identity(%onnx::Conv_716)
%onnx::Conv_722 = Identity(%onnx::Conv_719)
%onnx::Conv_713 = Identity(%onnx::Conv_710)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_706, %onnx::Conv_707)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.4/Add_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.5/relu/Relu_output_0 = Relu(%/cells.5/conv/Conv_output_0)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/relu/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/relu/Relu_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_823, %onnx::Conv_824)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_826, %onnx::Conv_827)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_829, %onnx::Conv_830)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_832, %onnx::Conv_833)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_835, %onnx::Conv_836)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_838, %onnx::Conv_839)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_841, %onnx::Conv_842)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_844, %onnx::Conv_845)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_847, %onnx::Conv_848)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_850, %onnx::Conv_851)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_853, %onnx::Conv_854)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_856, %onnx::Conv_857)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_output_0 = Reshape(%/cells.18/nl/Relu_output_0, %/cells.18/shuffle/Constant_output_0)
%/cells.18/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.18/shuffle/Reshape_output_0)
%/cells.18/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_1_output_0 = Reshape(%/cells.18/shuffle/Transpose_output_0, %/cells.18/shuffle/Constant_1_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/shuffle/Reshape_1_output_0, %onnx::Conv_859, %onnx::Conv_860)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_862, %onnx::Conv_863)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_865, %onnx::Conv_866)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_868, %onnx::Conv_869)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_871, %onnx::Conv_872)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_874, %onnx::Conv_875)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_877, %onnx::Conv_878)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_880, %onnx::Conv_881)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_883, %onnx::Conv_884)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_886, %onnx::Conv_887)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_889, %onnx::Conv_890)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_892, %onnx::Conv_893)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%704 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %704
}
|
val_accuracy
| 0
| 73,385,600
| 1,737,604
|
{'zcp_synflow': 76.09937568758984, 'zcp_zen': 69.43246459960938, 'zcp_epe_nas': 16.065749882262807, 'zcp_fisher': 0.12873536348342896, 'zcp_flops': 73385600.0, 'zcp_grad_norm': 27.238536834716797, 'zcp_grasp': -0.047557830810546875, 'zcp_jacov': -16.053498886545203, 'zcp_l2_norm': 618.617919921875, 'zcp_nwot': 214.93211931806866, 'zcp_params': 1737604.0, 'zcp_plain': 0.003151958342641592, 'zcp_snip': 43.429264068603516, 'lat_1080ti_1': 0.6120841557045352, 'lat_1080ti_32': 0.7580131332285059, 'lat_1080ti_64': 0.6134207946011124, 'lat_2080ti_1': 0.7081203517285688, 'lat_2080ti_32': 0.73955477308132, 'lat_2080ti_64': 0.6453758310359896, 'lat_essential_ph_1': 0.33962264150943394, 'lat_eyeriss': 0.5301743898190087, 'lat_fpga': 0.47499975261975685, 'lat_gold_6226': 0.3491083739653066, 'lat_gold_6240': 0.5390954727273893, 'lat_pixel2': 0.4782608695652174, 'lat_pixel3': 0.5388466702823381, 'lat_raspi4': 0.549653246430215, 'lat_samsung_a50': 0.3157894736842105, 'lat_samsung_s7': 0.1889763779527559, 'lat_silver_4114': 0.5547885616855295, 'lat_silver_4210r': 0.5739611163327426, 'lat_titan_rtx_1': 0.6708923143511674, 'lat_titan_rtx_32': 0.707148849324352, 'lat_titan_rtx_64': 0.6780412216131417, 'lat_titanx_1': 0.35268149427613954, 'lat_titanx_32': 0.6893783225077861, 'lat_titanx_64': 0.6603779477008297, 'lat_titanxp_1': 0.6373720488940692, 'lat_titanxp_32': 0.7014413716031378, 'lat_titanxp_64': 0.6396732075587694}
| |
FBNet_3694
|
FBNet
|
3694
|
3694
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_605[FLOAT, 16x3x3x3]
%onnx::Conv_606[FLOAT, 16]
%onnx::Conv_608[FLOAT, 48x16x1x1]
%onnx::Conv_609[FLOAT, 48]
%onnx::Conv_611[FLOAT, 48x1x5x5]
%onnx::Conv_614[FLOAT, 24x48x1x1]
%onnx::Conv_615[FLOAT, 24]
%onnx::Conv_617[FLOAT, 24x24x1x1]
%onnx::Conv_620[FLOAT, 24x1x5x5]
%onnx::Conv_623[FLOAT, 24x24x1x1]
%onnx::Conv_626[FLOAT, 72x24x1x1]
%onnx::Conv_627[FLOAT, 72]
%onnx::Conv_629[FLOAT, 72x1x5x5]
%onnx::Conv_632[FLOAT, 24x72x1x1]
%onnx::Conv_635[FLOAT, 72x24x1x1]
%onnx::Conv_638[FLOAT, 72x1x5x5]
%onnx::Conv_641[FLOAT, 24x72x1x1]
%onnx::Conv_644[FLOAT, 24x24x1x1]
%onnx::Conv_647[FLOAT, 24x1x3x3]
%onnx::Conv_650[FLOAT, 32x24x1x1]
%onnx::Conv_651[FLOAT, 32]
%onnx::Conv_653[FLOAT, 32x16x1x1]
%onnx::Conv_656[FLOAT, 32x1x5x5]
%onnx::Conv_659[FLOAT, 32x16x1x1]
%onnx::Conv_662[FLOAT, 32x16x1x1]
%onnx::Conv_665[FLOAT, 32x1x5x5]
%onnx::Conv_668[FLOAT, 32x16x1x1]
%onnx::Conv_671[FLOAT, 96x32x1x1]
%onnx::Conv_672[FLOAT, 96]
%onnx::Conv_674[FLOAT, 96x1x3x3]
%onnx::Conv_677[FLOAT, 32x96x1x1]
%onnx::Conv_680[FLOAT, 32x32x1x1]
%onnx::Conv_683[FLOAT, 32x1x3x3]
%onnx::Conv_686[FLOAT, 64x32x1x1]
%onnx::Conv_687[FLOAT, 64]
%onnx::Conv_689[FLOAT, 192x64x1x1]
%onnx::Conv_690[FLOAT, 192]
%onnx::Conv_692[FLOAT, 192x1x3x3]
%onnx::Conv_695[FLOAT, 64x192x1x1]
%onnx::Conv_698[FLOAT, 384x64x1x1]
%onnx::Conv_699[FLOAT, 384]
%onnx::Conv_701[FLOAT, 384x1x3x3]
%onnx::Conv_704[FLOAT, 64x384x1x1]
%onnx::Conv_707[FLOAT, 64x64x1x1]
%onnx::Conv_710[FLOAT, 64x1x3x3]
%onnx::Conv_713[FLOAT, 64x64x1x1]
%onnx::Conv_716[FLOAT, 112x64x1x1]
%onnx::Conv_717[FLOAT, 112]
%onnx::Conv_719[FLOAT, 112x112x1x1]
%onnx::Conv_722[FLOAT, 112x1x5x5]
%onnx::Conv_725[FLOAT, 112x112x1x1]
%onnx::Conv_728[FLOAT, 112x112x1x1]
%onnx::Conv_731[FLOAT, 112x1x3x3]
%onnx::Conv_734[FLOAT, 112x112x1x1]
%onnx::Conv_737[FLOAT, 672x112x1x1]
%onnx::Conv_738[FLOAT, 672]
%onnx::Conv_740[FLOAT, 672x1x5x5]
%onnx::Conv_743[FLOAT, 184x672x1x1]
%onnx::Conv_744[FLOAT, 184]
%onnx::Conv_746[FLOAT, 1104x184x1x1]
%onnx::Conv_747[FLOAT, 1104]
%onnx::Conv_749[FLOAT, 1104x1x3x3]
%onnx::Conv_752[FLOAT, 184x1104x1x1]
%onnx::Conv_755[FLOAT, 184x92x1x1]
%onnx::Conv_758[FLOAT, 184x1x5x5]
%onnx::Conv_761[FLOAT, 184x92x1x1]
%onnx::Conv_764[FLOAT, 184x92x1x1]
%onnx::Conv_767[FLOAT, 184x1x3x3]
%onnx::Conv_770[FLOAT, 184x92x1x1]
%onnx::Conv_773[FLOAT, 352x184x1x1]
%onnx::Conv_774[FLOAT, 352]
%onnx::Conv_776[FLOAT, 1504x352x1x1]
%onnx::Conv_777[FLOAT, 1504]
) {
%onnx::Conv_771 = Identity(%onnx::Conv_744)
%onnx::Conv_768 = Identity(%onnx::Conv_744)
%onnx::Conv_765 = Identity(%onnx::Conv_744)
%onnx::Conv_762 = Identity(%onnx::Conv_744)
%onnx::Conv_759 = Identity(%onnx::Conv_744)
%onnx::Conv_756 = Identity(%onnx::Conv_744)
%onnx::Conv_753 = Identity(%onnx::Conv_744)
%onnx::Conv_750 = Identity(%onnx::Conv_747)
%onnx::Conv_741 = Identity(%onnx::Conv_738)
%onnx::Conv_735 = Identity(%onnx::Conv_717)
%onnx::Conv_732 = Identity(%onnx::Conv_717)
%onnx::Conv_729 = Identity(%onnx::Conv_717)
%onnx::Conv_726 = Identity(%onnx::Conv_717)
%onnx::Conv_723 = Identity(%onnx::Conv_717)
%onnx::Conv_720 = Identity(%onnx::Conv_717)
%onnx::Conv_714 = Identity(%onnx::Conv_687)
%onnx::Conv_711 = Identity(%onnx::Conv_687)
%onnx::Conv_708 = Identity(%onnx::Conv_687)
%onnx::Conv_705 = Identity(%onnx::Conv_687)
%onnx::Conv_702 = Identity(%onnx::Conv_699)
%onnx::Conv_696 = Identity(%onnx::Conv_687)
%onnx::Conv_693 = Identity(%onnx::Conv_690)
%onnx::Conv_684 = Identity(%onnx::Conv_651)
%onnx::Conv_681 = Identity(%onnx::Conv_651)
%onnx::Conv_678 = Identity(%onnx::Conv_651)
%onnx::Conv_675 = Identity(%onnx::Conv_672)
%onnx::Conv_669 = Identity(%onnx::Conv_651)
%onnx::Conv_666 = Identity(%onnx::Conv_651)
%onnx::Conv_663 = Identity(%onnx::Conv_651)
%onnx::Conv_660 = Identity(%onnx::Conv_651)
%onnx::Conv_657 = Identity(%onnx::Conv_651)
%onnx::Conv_654 = Identity(%onnx::Conv_651)
%onnx::Conv_648 = Identity(%onnx::Conv_615)
%onnx::Conv_645 = Identity(%onnx::Conv_615)
%onnx::Conv_642 = Identity(%onnx::Conv_615)
%onnx::Conv_639 = Identity(%onnx::Conv_627)
%onnx::Conv_636 = Identity(%onnx::Conv_627)
%onnx::Conv_633 = Identity(%onnx::Conv_615)
%onnx::Conv_630 = Identity(%onnx::Conv_627)
%onnx::Conv_624 = Identity(%onnx::Conv_615)
%onnx::Conv_621 = Identity(%onnx::Conv_615)
%onnx::Conv_618 = Identity(%onnx::Conv_615)
%onnx::Conv_612 = Identity(%onnx::Conv_609)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_605, %onnx::Conv_606)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_608, %onnx::Conv_609)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_611, %onnx::Conv_612)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_614, %onnx::Conv_615)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_617, %onnx::Conv_618)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_620, %onnx::Conv_621)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_623, %onnx::Conv_624)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.13/relu/Relu_output_0 = Relu(%/cells.13/conv/Conv_output_0)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/relu/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/relu/Relu_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.21/relu/Relu_output_0 = Relu(%/cells.21/conv/Conv_output_0)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/relu/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%603 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %603
}
|
val_accuracy
| 0
| 56,710,528
| 1,651,396
|
{'zcp_synflow': 74.75443375555011, 'zcp_zen': 62.603607177734375, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.07546202093362808, 'zcp_flops': 56710528.0, 'zcp_grad_norm': 17.111940383911133, 'zcp_grasp': -0.019321441650390625, 'zcp_jacov': -16.06563820635826, 'zcp_l2_norm': 552.5843505859375, 'zcp_nwot': 208.60511478961212, 'zcp_params': 1651396.0, 'zcp_plain': -0.0003486708737909794, 'zcp_snip': 30.07236099243164, 'lat_1080ti_1': 0.4746677467547743, 'lat_1080ti_32': 0.46064638659095386, 'lat_1080ti_64': 0.3555105753504757, 'lat_2080ti_1': 0.482672083161811, 'lat_2080ti_32': 0.4017286031201697, 'lat_2080ti_64': 0.33295959513217255, 'lat_essential_ph_1': 0.22641509433962265, 'lat_eyeriss': 0.309885757247905, 'lat_fpga': 0.2483400785679652, 'lat_gold_6226': 0.23980554464149212, 'lat_gold_6240': 0.360813350879894, 'lat_pixel2': 0.21739130434782608, 'lat_pixel3': 0.31263166355762423, 'lat_raspi4': 0.3129957921669798, 'lat_samsung_a50': 0.11578947368421053, 'lat_samsung_s7': 0.11023622047244094, 'lat_silver_4114': 0.3792382213854042, 'lat_silver_4210r': 0.39820397085147374, 'lat_titan_rtx_1': 0.4447912995963583, 'lat_titan_rtx_32': 0.39866561913360565, 'lat_titan_rtx_64': 0.3485891313671661, 'lat_titanx_1': 0.23527763220810227, 'lat_titanx_32': 0.3691293846410788, 'lat_titanx_64': 0.34948660275273546, 'lat_titanxp_1': 0.4262585978406213, 'lat_titanxp_32': 0.41179943351771375, 'lat_titanxp_64': 0.3457745785381857}
| |
FBNet_4571
|
FBNet
|
4571
|
4571
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_640[FLOAT, 16x3x3x3]
%onnx::Conv_641[FLOAT, 16]
%onnx::Conv_643[FLOAT, 96x16x1x1]
%onnx::Conv_644[FLOAT, 96]
%onnx::Conv_646[FLOAT, 96x1x5x5]
%onnx::Conv_649[FLOAT, 16x96x1x1]
%onnx::Conv_652[FLOAT, 48x16x1x1]
%onnx::Conv_653[FLOAT, 48]
%onnx::Conv_655[FLOAT, 48x1x5x5]
%onnx::Conv_658[FLOAT, 24x48x1x1]
%onnx::Conv_659[FLOAT, 24]
%onnx::Conv_661[FLOAT, 72x24x1x1]
%onnx::Conv_662[FLOAT, 72]
%onnx::Conv_664[FLOAT, 72x1x5x5]
%onnx::Conv_667[FLOAT, 24x72x1x1]
%onnx::Conv_670[FLOAT, 24x12x1x1]
%onnx::Conv_673[FLOAT, 24x1x3x3]
%onnx::Conv_676[FLOAT, 24x12x1x1]
%onnx::Conv_679[FLOAT, 144x24x1x1]
%onnx::Conv_680[FLOAT, 144]
%onnx::Conv_682[FLOAT, 144x1x5x5]
%onnx::Conv_685[FLOAT, 24x144x1x1]
%onnx::Conv_688[FLOAT, 72x24x1x1]
%onnx::Conv_691[FLOAT, 72x1x5x5]
%onnx::Conv_694[FLOAT, 32x72x1x1]
%onnx::Conv_695[FLOAT, 32]
%onnx::Conv_697[FLOAT, 192x32x1x1]
%onnx::Conv_698[FLOAT, 192]
%onnx::Conv_700[FLOAT, 192x1x3x3]
%onnx::Conv_703[FLOAT, 32x192x1x1]
%onnx::Conv_706[FLOAT, 96x32x1x1]
%onnx::Conv_709[FLOAT, 96x1x5x5]
%onnx::Conv_712[FLOAT, 32x96x1x1]
%onnx::Conv_715[FLOAT, 96x32x1x1]
%onnx::Conv_718[FLOAT, 96x1x5x5]
%onnx::Conv_721[FLOAT, 32x96x1x1]
%onnx::Conv_724[FLOAT, 192x32x1x1]
%onnx::Conv_727[FLOAT, 192x1x3x3]
%onnx::Conv_730[FLOAT, 64x192x1x1]
%onnx::Conv_731[FLOAT, 64]
%onnx::Conv_733[FLOAT, 192x64x1x1]
%onnx::Conv_736[FLOAT, 192x1x3x3]
%onnx::Conv_739[FLOAT, 64x192x1x1]
%onnx::Conv_742[FLOAT, 384x64x1x1]
%onnx::Conv_743[FLOAT, 384]
%onnx::Conv_745[FLOAT, 384x1x5x5]
%onnx::Conv_748[FLOAT, 64x384x1x1]
%onnx::Conv_751[FLOAT, 192x64x1x1]
%onnx::Conv_754[FLOAT, 192x1x3x3]
%onnx::Conv_757[FLOAT, 64x192x1x1]
%onnx::Conv_760[FLOAT, 112x64x1x1]
%onnx::Conv_761[FLOAT, 112]
%onnx::Conv_763[FLOAT, 112x112x1x1]
%onnx::Conv_766[FLOAT, 112x1x5x5]
%onnx::Conv_769[FLOAT, 112x112x1x1]
%onnx::Conv_772[FLOAT, 672x112x1x1]
%onnx::Conv_773[FLOAT, 672]
%onnx::Conv_775[FLOAT, 672x1x3x3]
%onnx::Conv_778[FLOAT, 112x672x1x1]
%onnx::Conv_781[FLOAT, 112x56x1x1]
%onnx::Conv_784[FLOAT, 112x1x3x3]
%onnx::Conv_787[FLOAT, 112x56x1x1]
%onnx::Conv_790[FLOAT, 672x112x1x1]
%onnx::Conv_793[FLOAT, 672x1x3x3]
%onnx::Conv_796[FLOAT, 184x672x1x1]
%onnx::Conv_797[FLOAT, 184]
%onnx::Conv_799[FLOAT, 184x184x1x1]
%onnx::Conv_802[FLOAT, 184x1x3x3]
%onnx::Conv_805[FLOAT, 184x184x1x1]
%onnx::Conv_808[FLOAT, 184x184x1x1]
%onnx::Conv_811[FLOAT, 184x1x3x3]
%onnx::Conv_814[FLOAT, 184x184x1x1]
%onnx::Conv_817[FLOAT, 552x184x1x1]
%onnx::Conv_818[FLOAT, 552]
%onnx::Conv_820[FLOAT, 552x1x5x5]
%onnx::Conv_823[FLOAT, 184x552x1x1]
%onnx::Conv_826[FLOAT, 1104x184x1x1]
%onnx::Conv_827[FLOAT, 1104]
%onnx::Conv_829[FLOAT, 1104x1x5x5]
%onnx::Conv_832[FLOAT, 352x1104x1x1]
%onnx::Conv_833[FLOAT, 352]
%onnx::Conv_835[FLOAT, 1504x352x1x1]
%onnx::Conv_836[FLOAT, 1504]
) {
%onnx::Conv_830 = Identity(%onnx::Conv_827)
%onnx::Conv_824 = Identity(%onnx::Conv_797)
%onnx::Conv_821 = Identity(%onnx::Conv_818)
%onnx::Conv_815 = Identity(%onnx::Conv_797)
%onnx::Conv_812 = Identity(%onnx::Conv_797)
%onnx::Conv_809 = Identity(%onnx::Conv_797)
%onnx::Conv_806 = Identity(%onnx::Conv_797)
%onnx::Conv_803 = Identity(%onnx::Conv_797)
%onnx::Conv_800 = Identity(%onnx::Conv_797)
%onnx::Conv_794 = Identity(%onnx::Conv_773)
%onnx::Conv_791 = Identity(%onnx::Conv_773)
%onnx::Conv_788 = Identity(%onnx::Conv_761)
%onnx::Conv_785 = Identity(%onnx::Conv_761)
%onnx::Conv_782 = Identity(%onnx::Conv_761)
%onnx::Conv_779 = Identity(%onnx::Conv_761)
%onnx::Conv_776 = Identity(%onnx::Conv_773)
%onnx::Conv_770 = Identity(%onnx::Conv_761)
%onnx::Conv_767 = Identity(%onnx::Conv_761)
%onnx::Conv_764 = Identity(%onnx::Conv_761)
%onnx::Conv_758 = Identity(%onnx::Conv_731)
%onnx::Conv_755 = Identity(%onnx::Conv_698)
%onnx::Conv_752 = Identity(%onnx::Conv_698)
%onnx::Conv_749 = Identity(%onnx::Conv_731)
%onnx::Conv_746 = Identity(%onnx::Conv_743)
%onnx::Conv_740 = Identity(%onnx::Conv_731)
%onnx::Conv_737 = Identity(%onnx::Conv_698)
%onnx::Conv_734 = Identity(%onnx::Conv_698)
%onnx::Conv_728 = Identity(%onnx::Conv_698)
%onnx::Conv_725 = Identity(%onnx::Conv_698)
%onnx::Conv_722 = Identity(%onnx::Conv_695)
%onnx::Conv_719 = Identity(%onnx::Conv_644)
%onnx::Conv_716 = Identity(%onnx::Conv_644)
%onnx::Conv_713 = Identity(%onnx::Conv_695)
%onnx::Conv_710 = Identity(%onnx::Conv_644)
%onnx::Conv_707 = Identity(%onnx::Conv_644)
%onnx::Conv_704 = Identity(%onnx::Conv_695)
%onnx::Conv_701 = Identity(%onnx::Conv_698)
%onnx::Conv_692 = Identity(%onnx::Conv_662)
%onnx::Conv_689 = Identity(%onnx::Conv_662)
%onnx::Conv_686 = Identity(%onnx::Conv_659)
%onnx::Conv_683 = Identity(%onnx::Conv_680)
%onnx::Conv_677 = Identity(%onnx::Conv_659)
%onnx::Conv_674 = Identity(%onnx::Conv_659)
%onnx::Conv_671 = Identity(%onnx::Conv_659)
%onnx::Conv_668 = Identity(%onnx::Conv_659)
%onnx::Conv_665 = Identity(%onnx::Conv_662)
%onnx::Conv_656 = Identity(%onnx::Conv_653)
%onnx::Conv_650 = Identity(%onnx::Conv_641)
%onnx::Conv_647 = Identity(%onnx::Conv_644)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_640, %onnx::Conv_641)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_643, %onnx::Conv_644)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_646, %onnx::Conv_647)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_649, %onnx::Conv_650)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_652, %onnx::Conv_653)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_655, %onnx::Conv_656)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_658, %onnx::Conv_659)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.13/relu/Relu_output_0 = Relu(%/cells.13/conv/Conv_output_0)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/relu/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/relu/Relu_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_823, %onnx::Conv_824)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_826, %onnx::Conv_827)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_829, %onnx::Conv_830)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_832, %onnx::Conv_833)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_835, %onnx::Conv_836)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%638 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %638
}
|
val_accuracy
| 0
| 95,003,264
| 2,287,780
|
{'zcp_synflow': 85.87723344166349, 'zcp_zen': 76.34773254394531, 'zcp_epe_nas': 21.235971972286617, 'zcp_fisher': 0.23431219160556793, 'zcp_flops': 95003264.0, 'zcp_grad_norm': 35.03923034667969, 'zcp_grasp': 0.115020751953125, 'zcp_jacov': -16.069284826002423, 'zcp_l2_norm': 727.5813598632812, 'zcp_nwot': 218.32900832501713, 'zcp_params': 2287780.0, 'zcp_plain': 0.0011953659122809768, 'zcp_snip': 61.422523498535156, 'lat_1080ti_1': 0.7617964543481734, 'lat_1080ti_32': 0.7992948468306427, 'lat_1080ti_64': 0.7267735477524503, 'lat_2080ti_1': 0.7129892962684266, 'lat_2080ti_32': 0.7131886845722701, 'lat_2080ti_64': 0.7235346628247638, 'lat_essential_ph_1': 0.4716981132075472, 'lat_eyeriss': 0.7947664716776996, 'lat_fpga': 0.7271296965139177, 'lat_gold_6226': 0.5352686741092134, 'lat_gold_6240': 0.673988569665993, 'lat_pixel2': 0.4782608695652174, 'lat_pixel3': 0.8038708655846415, 'lat_raspi4': 0.8130928224548688, 'lat_samsung_a50': 0.30526315789473685, 'lat_samsung_s7': 0.25196850393700787, 'lat_silver_4114': 0.65868719287173, 'lat_silver_4210r': 0.7111411991873431, 'lat_titan_rtx_1': 0.6701550201158116, 'lat_titan_rtx_32': 0.684757212812332, 'lat_titan_rtx_64': 0.7223886987317455, 'lat_titanx_1': 0.3572855645552966, 'lat_titanx_32': 0.7458163117781182, 'lat_titanx_64': 0.7884936226642563, 'lat_titanxp_1': 0.6399730821832554, 'lat_titanxp_32': 0.7538448511548521, 'lat_titanxp_64': 0.723976686755151}
| |
FBNet_4329
|
FBNet
|
4329
|
4329
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_723[FLOAT, 16x3x3x3]
%onnx::Conv_724[FLOAT, 16]
%onnx::Conv_726[FLOAT, 16x8x1x1]
%onnx::Conv_729[FLOAT, 16x1x5x5]
%onnx::Conv_732[FLOAT, 16x8x1x1]
%onnx::Conv_735[FLOAT, 16x8x1x1]
%onnx::Conv_738[FLOAT, 16x1x3x3]
%onnx::Conv_741[FLOAT, 24x8x1x1]
%onnx::Conv_742[FLOAT, 24]
%onnx::Conv_744[FLOAT, 24x24x1x1]
%onnx::Conv_747[FLOAT, 24x1x3x3]
%onnx::Conv_750[FLOAT, 24x24x1x1]
%onnx::Conv_753[FLOAT, 24x24x1x1]
%onnx::Conv_756[FLOAT, 24x1x5x5]
%onnx::Conv_759[FLOAT, 24x24x1x1]
%onnx::Conv_762[FLOAT, 24x12x1x1]
%onnx::Conv_765[FLOAT, 24x1x3x3]
%onnx::Conv_768[FLOAT, 24x12x1x1]
%onnx::Conv_771[FLOAT, 24x24x1x1]
%onnx::Conv_774[FLOAT, 24x1x5x5]
%onnx::Conv_777[FLOAT, 32x24x1x1]
%onnx::Conv_778[FLOAT, 32]
%onnx::Conv_780[FLOAT, 96x32x1x1]
%onnx::Conv_781[FLOAT, 96]
%onnx::Conv_783[FLOAT, 96x1x5x5]
%onnx::Conv_786[FLOAT, 32x96x1x1]
%onnx::Conv_789[FLOAT, 32x32x1x1]
%onnx::Conv_792[FLOAT, 32x1x5x5]
%onnx::Conv_795[FLOAT, 32x32x1x1]
%onnx::Conv_798[FLOAT, 32x16x1x1]
%onnx::Conv_801[FLOAT, 32x1x3x3]
%onnx::Conv_804[FLOAT, 32x16x1x1]
%onnx::Conv_807[FLOAT, 192x32x1x1]
%onnx::Conv_808[FLOAT, 192]
%onnx::Conv_810[FLOAT, 192x1x5x5]
%onnx::Conv_813[FLOAT, 64x192x1x1]
%onnx::Conv_814[FLOAT, 64]
%onnx::Conv_816[FLOAT, 384x64x1x1]
%onnx::Conv_817[FLOAT, 384]
%onnx::Conv_819[FLOAT, 384x1x3x3]
%onnx::Conv_822[FLOAT, 64x384x1x1]
%onnx::Conv_825[FLOAT, 384x64x1x1]
%onnx::Conv_828[FLOAT, 384x1x5x5]
%onnx::Conv_831[FLOAT, 64x384x1x1]
%onnx::Conv_834[FLOAT, 64x32x1x1]
%onnx::Conv_837[FLOAT, 64x1x5x5]
%onnx::Conv_840[FLOAT, 64x32x1x1]
%onnx::Conv_843[FLOAT, 64x32x1x1]
%onnx::Conv_846[FLOAT, 64x1x3x3]
%onnx::Conv_849[FLOAT, 112x32x1x1]
%onnx::Conv_850[FLOAT, 112]
%onnx::Conv_852[FLOAT, 672x112x1x1]
%onnx::Conv_853[FLOAT, 672]
%onnx::Conv_855[FLOAT, 672x1x3x3]
%onnx::Conv_858[FLOAT, 112x672x1x1]
%onnx::Conv_861[FLOAT, 672x112x1x1]
%onnx::Conv_864[FLOAT, 672x1x3x3]
%onnx::Conv_867[FLOAT, 112x672x1x1]
%onnx::Conv_870[FLOAT, 336x112x1x1]
%onnx::Conv_871[FLOAT, 336]
%onnx::Conv_873[FLOAT, 336x1x5x5]
%onnx::Conv_876[FLOAT, 184x336x1x1]
%onnx::Conv_877[FLOAT, 184]
%onnx::Conv_879[FLOAT, 184x184x1x1]
%onnx::Conv_882[FLOAT, 184x1x5x5]
%onnx::Conv_885[FLOAT, 184x184x1x1]
%onnx::Conv_888[FLOAT, 1104x184x1x1]
%onnx::Conv_889[FLOAT, 1104]
%onnx::Conv_891[FLOAT, 1104x1x5x5]
%onnx::Conv_894[FLOAT, 184x1104x1x1]
%onnx::Conv_897[FLOAT, 1104x184x1x1]
%onnx::Conv_900[FLOAT, 1104x1x3x3]
%onnx::Conv_903[FLOAT, 184x1104x1x1]
%onnx::Conv_906[FLOAT, 184x92x1x1]
%onnx::Conv_909[FLOAT, 184x1x5x5]
%onnx::Conv_912[FLOAT, 352x92x1x1]
%onnx::Conv_913[FLOAT, 352]
%onnx::Conv_915[FLOAT, 1504x352x1x1]
%onnx::Conv_916[FLOAT, 1504]
) {
%onnx::Conv_910 = Identity(%onnx::Conv_877)
%onnx::Conv_907 = Identity(%onnx::Conv_877)
%onnx::Conv_904 = Identity(%onnx::Conv_877)
%onnx::Conv_901 = Identity(%onnx::Conv_889)
%onnx::Conv_898 = Identity(%onnx::Conv_889)
%onnx::Conv_895 = Identity(%onnx::Conv_877)
%onnx::Conv_892 = Identity(%onnx::Conv_889)
%onnx::Conv_886 = Identity(%onnx::Conv_877)
%onnx::Conv_883 = Identity(%onnx::Conv_877)
%onnx::Conv_880 = Identity(%onnx::Conv_877)
%onnx::Conv_874 = Identity(%onnx::Conv_871)
%onnx::Conv_868 = Identity(%onnx::Conv_850)
%onnx::Conv_865 = Identity(%onnx::Conv_853)
%onnx::Conv_862 = Identity(%onnx::Conv_853)
%onnx::Conv_859 = Identity(%onnx::Conv_850)
%onnx::Conv_856 = Identity(%onnx::Conv_853)
%onnx::Conv_847 = Identity(%onnx::Conv_814)
%onnx::Conv_844 = Identity(%onnx::Conv_814)
%onnx::Conv_841 = Identity(%onnx::Conv_814)
%onnx::Conv_838 = Identity(%onnx::Conv_814)
%onnx::Conv_835 = Identity(%onnx::Conv_814)
%onnx::Conv_832 = Identity(%onnx::Conv_814)
%onnx::Conv_829 = Identity(%onnx::Conv_817)
%onnx::Conv_826 = Identity(%onnx::Conv_817)
%onnx::Conv_823 = Identity(%onnx::Conv_814)
%onnx::Conv_820 = Identity(%onnx::Conv_817)
%onnx::Conv_811 = Identity(%onnx::Conv_808)
%onnx::Conv_805 = Identity(%onnx::Conv_778)
%onnx::Conv_802 = Identity(%onnx::Conv_778)
%onnx::Conv_799 = Identity(%onnx::Conv_778)
%onnx::Conv_796 = Identity(%onnx::Conv_778)
%onnx::Conv_793 = Identity(%onnx::Conv_778)
%onnx::Conv_790 = Identity(%onnx::Conv_778)
%onnx::Conv_787 = Identity(%onnx::Conv_778)
%onnx::Conv_784 = Identity(%onnx::Conv_781)
%onnx::Conv_775 = Identity(%onnx::Conv_742)
%onnx::Conv_772 = Identity(%onnx::Conv_742)
%onnx::Conv_769 = Identity(%onnx::Conv_742)
%onnx::Conv_766 = Identity(%onnx::Conv_742)
%onnx::Conv_763 = Identity(%onnx::Conv_742)
%onnx::Conv_760 = Identity(%onnx::Conv_742)
%onnx::Conv_757 = Identity(%onnx::Conv_742)
%onnx::Conv_754 = Identity(%onnx::Conv_742)
%onnx::Conv_751 = Identity(%onnx::Conv_742)
%onnx::Conv_748 = Identity(%onnx::Conv_742)
%onnx::Conv_745 = Identity(%onnx::Conv_742)
%onnx::Conv_739 = Identity(%onnx::Conv_724)
%onnx::Conv_736 = Identity(%onnx::Conv_724)
%onnx::Conv_733 = Identity(%onnx::Conv_724)
%onnx::Conv_730 = Identity(%onnx::Conv_724)
%onnx::Conv_727 = Identity(%onnx::Conv_724)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_723, %onnx::Conv_724)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_858, %onnx::Conv_859)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_861, %onnx::Conv_862)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_864, %onnx::Conv_865)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_867, %onnx::Conv_868)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_870, %onnx::Conv_871)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_873, %onnx::Conv_874)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_876, %onnx::Conv_877)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_879, %onnx::Conv_880)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_882, %onnx::Conv_883)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_885, %onnx::Conv_886)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_888, %onnx::Conv_889)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_891, %onnx::Conv_892)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_894, %onnx::Conv_895)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_897, %onnx::Conv_898)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_900, %onnx::Conv_901)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_903, %onnx::Conv_904)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_906, %onnx::Conv_907)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_909, %onnx::Conv_910)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_912, %onnx::Conv_913)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_915, %onnx::Conv_916)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%721 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %721
}
|
val_accuracy
| 0
| 70,631,296
| 2,273,236
|
{'zcp_synflow': 76.74998819373809, 'zcp_zen': 69.07591247558594, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.12420324981212616, 'zcp_flops': 70631296.0, 'zcp_grad_norm': 24.546001434326172, 'zcp_grasp': 0.00125885009765625, 'zcp_jacov': -16.061292745686266, 'zcp_l2_norm': 654.7416381835938, 'zcp_nwot': 208.25829708311744, 'zcp_params': 2273236.0, 'zcp_plain': 0.010439128614962101, 'zcp_snip': 41.53748321533203, 'lat_1080ti_1': 0.6824222980547053, 'lat_1080ti_32': 0.5602014802605711, 'lat_1080ti_64': 0.3368597720344995, 'lat_2080ti_1': 0.7687564837970476, 'lat_2080ti_32': 0.5935816816687653, 'lat_2080ti_64': 0.3362049147774844, 'lat_essential_ph_1': 0.37735849056603776, 'lat_eyeriss': 0.4357820035574789, 'lat_fpga': 0.5242680018602993, 'lat_gold_6226': 0.4957792195316473, 'lat_gold_6240': 0.7410943568307263, 'lat_pixel2': 0.5217391304347826, 'lat_pixel3': 0.42270847232155273, 'lat_raspi4': 0.48027534373661374, 'lat_samsung_a50': 0.24210526315789474, 'lat_samsung_s7': 0.1732283464566929, 'lat_silver_4114': 0.774156300883419, 'lat_silver_4210r': 0.8465507429049344, 'lat_titan_rtx_1': 0.7328848127961719, 'lat_titan_rtx_32': 0.6010768561992706, 'lat_titan_rtx_64': 0.3855235864236182, 'lat_titanx_1': 0.40608914403922614, 'lat_titanx_32': 0.46273069439003117, 'lat_titanx_64': 0.3129100330135949, 'lat_titanxp_1': 0.6988637751815387, 'lat_titanxp_32': 0.5355753713224457, 'lat_titanxp_64': 0.330627475807337}
| |
FBNet_353
|
FBNet
|
353
|
353
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_616[FLOAT, 16x3x3x3]
%onnx::Conv_617[FLOAT, 16]
%onnx::Conv_619[FLOAT, 16x16x1x1]
%onnx::Conv_622[FLOAT, 16x1x3x3]
%onnx::Conv_625[FLOAT, 16x16x1x1]
%onnx::Conv_628[FLOAT, 16x16x1x1]
%onnx::Conv_631[FLOAT, 16x1x3x3]
%onnx::Conv_634[FLOAT, 24x16x1x1]
%onnx::Conv_635[FLOAT, 24]
%onnx::Conv_637[FLOAT, 72x24x1x1]
%onnx::Conv_638[FLOAT, 72]
%onnx::Conv_640[FLOAT, 72x1x3x3]
%onnx::Conv_643[FLOAT, 24x72x1x1]
%onnx::Conv_646[FLOAT, 24x12x1x1]
%onnx::Conv_649[FLOAT, 24x1x5x5]
%onnx::Conv_652[FLOAT, 24x12x1x1]
%onnx::Conv_655[FLOAT, 24x24x1x1]
%onnx::Conv_658[FLOAT, 24x1x3x3]
%onnx::Conv_661[FLOAT, 32x24x1x1]
%onnx::Conv_662[FLOAT, 32]
%onnx::Conv_664[FLOAT, 192x32x1x1]
%onnx::Conv_665[FLOAT, 192]
%onnx::Conv_667[FLOAT, 192x1x5x5]
%onnx::Conv_670[FLOAT, 32x192x1x1]
%onnx::Conv_673[FLOAT, 96x32x1x1]
%onnx::Conv_674[FLOAT, 96]
%onnx::Conv_676[FLOAT, 96x1x3x3]
%onnx::Conv_679[FLOAT, 32x96x1x1]
%onnx::Conv_682[FLOAT, 32x16x1x1]
%onnx::Conv_685[FLOAT, 32x1x5x5]
%onnx::Conv_688[FLOAT, 64x16x1x1]
%onnx::Conv_689[FLOAT, 64]
%onnx::Conv_691[FLOAT, 64x64x1x1]
%onnx::Conv_694[FLOAT, 64x1x3x3]
%onnx::Conv_697[FLOAT, 64x64x1x1]
%onnx::Conv_700[FLOAT, 64x32x1x1]
%onnx::Conv_703[FLOAT, 64x1x3x3]
%onnx::Conv_706[FLOAT, 64x32x1x1]
%onnx::Conv_709[FLOAT, 112x64x1x1]
%onnx::Conv_710[FLOAT, 112]
%onnx::Conv_712[FLOAT, 112x112x1x1]
%onnx::Conv_715[FLOAT, 112x1x3x3]
%onnx::Conv_718[FLOAT, 112x112x1x1]
%onnx::Conv_721[FLOAT, 112x56x1x1]
%onnx::Conv_724[FLOAT, 112x1x5x5]
%onnx::Conv_727[FLOAT, 112x56x1x1]
%onnx::Conv_730[FLOAT, 336x112x1x1]
%onnx::Conv_731[FLOAT, 336]
%onnx::Conv_733[FLOAT, 336x1x5x5]
%onnx::Conv_736[FLOAT, 112x336x1x1]
%onnx::Conv_739[FLOAT, 336x112x1x1]
%onnx::Conv_742[FLOAT, 336x1x5x5]
%onnx::Conv_745[FLOAT, 184x336x1x1]
%onnx::Conv_746[FLOAT, 184]
%onnx::Conv_748[FLOAT, 184x184x1x1]
%onnx::Conv_751[FLOAT, 184x1x5x5]
%onnx::Conv_754[FLOAT, 184x184x1x1]
%onnx::Conv_757[FLOAT, 184x92x1x1]
%onnx::Conv_760[FLOAT, 184x1x5x5]
%onnx::Conv_763[FLOAT, 184x92x1x1]
%onnx::Conv_766[FLOAT, 184x184x1x1]
%onnx::Conv_769[FLOAT, 184x1x5x5]
%onnx::Conv_772[FLOAT, 184x184x1x1]
%onnx::Conv_775[FLOAT, 1104x184x1x1]
%onnx::Conv_776[FLOAT, 1104]
%onnx::Conv_778[FLOAT, 1104x1x3x3]
%onnx::Conv_781[FLOAT, 352x1104x1x1]
%onnx::Conv_782[FLOAT, 352]
%onnx::Conv_784[FLOAT, 1504x352x1x1]
%onnx::Conv_785[FLOAT, 1504]
) {
%onnx::Conv_779 = Identity(%onnx::Conv_776)
%onnx::Conv_773 = Identity(%onnx::Conv_746)
%onnx::Conv_770 = Identity(%onnx::Conv_746)
%onnx::Conv_767 = Identity(%onnx::Conv_746)
%onnx::Conv_764 = Identity(%onnx::Conv_746)
%onnx::Conv_761 = Identity(%onnx::Conv_746)
%onnx::Conv_758 = Identity(%onnx::Conv_746)
%onnx::Conv_755 = Identity(%onnx::Conv_746)
%onnx::Conv_752 = Identity(%onnx::Conv_746)
%onnx::Conv_749 = Identity(%onnx::Conv_746)
%onnx::Conv_743 = Identity(%onnx::Conv_731)
%onnx::Conv_740 = Identity(%onnx::Conv_731)
%onnx::Conv_737 = Identity(%onnx::Conv_710)
%onnx::Conv_734 = Identity(%onnx::Conv_731)
%onnx::Conv_728 = Identity(%onnx::Conv_710)
%onnx::Conv_725 = Identity(%onnx::Conv_710)
%onnx::Conv_722 = Identity(%onnx::Conv_710)
%onnx::Conv_719 = Identity(%onnx::Conv_710)
%onnx::Conv_716 = Identity(%onnx::Conv_710)
%onnx::Conv_713 = Identity(%onnx::Conv_710)
%onnx::Conv_707 = Identity(%onnx::Conv_689)
%onnx::Conv_704 = Identity(%onnx::Conv_689)
%onnx::Conv_701 = Identity(%onnx::Conv_689)
%onnx::Conv_698 = Identity(%onnx::Conv_689)
%onnx::Conv_695 = Identity(%onnx::Conv_689)
%onnx::Conv_692 = Identity(%onnx::Conv_689)
%onnx::Conv_686 = Identity(%onnx::Conv_662)
%onnx::Conv_683 = Identity(%onnx::Conv_662)
%onnx::Conv_680 = Identity(%onnx::Conv_662)
%onnx::Conv_677 = Identity(%onnx::Conv_674)
%onnx::Conv_671 = Identity(%onnx::Conv_662)
%onnx::Conv_668 = Identity(%onnx::Conv_665)
%onnx::Conv_659 = Identity(%onnx::Conv_635)
%onnx::Conv_656 = Identity(%onnx::Conv_635)
%onnx::Conv_653 = Identity(%onnx::Conv_635)
%onnx::Conv_650 = Identity(%onnx::Conv_635)
%onnx::Conv_647 = Identity(%onnx::Conv_635)
%onnx::Conv_644 = Identity(%onnx::Conv_635)
%onnx::Conv_641 = Identity(%onnx::Conv_638)
%onnx::Conv_632 = Identity(%onnx::Conv_617)
%onnx::Conv_629 = Identity(%onnx::Conv_617)
%onnx::Conv_626 = Identity(%onnx::Conv_617)
%onnx::Conv_623 = Identity(%onnx::Conv_617)
%onnx::Conv_620 = Identity(%onnx::Conv_617)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_616, %onnx::Conv_617)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_619, %onnx::Conv_620)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_622, %onnx::Conv_623)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_625, %onnx::Conv_626)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_628, %onnx::Conv_629)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_631, %onnx::Conv_632)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_634, %onnx::Conv_635)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_637, %onnx::Conv_638)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_640, %onnx::Conv_641)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_643, %onnx::Conv_644)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_646, %onnx::Conv_647)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_649, %onnx::Conv_650)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_652, %onnx::Conv_653)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_655, %onnx::Conv_656)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_658, %onnx::Conv_659)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.13/relu/Relu_output_0 = Relu(%/cells.13/conv/Conv_output_0)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/relu/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/relu/Relu_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%614 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %614
}
|
val_accuracy
| 0
| 50,641,024
| 1,772,852
|
{'zcp_synflow': 70.52842920654639, 'zcp_zen': 59.84412384033203, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.08622127026319504, 'zcp_flops': 50641024.0, 'zcp_grad_norm': 17.043773651123047, 'zcp_grasp': -0.017660140991210938, 'zcp_jacov': -16.073890019115957, 'zcp_l2_norm': 533.25634765625, 'zcp_nwot': 205.04157142114843, 'zcp_params': 1772852.0, 'zcp_plain': 0.005153720732778311, 'zcp_snip': 29.840438842773438, 'lat_1080ti_1': 0.3709892956270388, 'lat_1080ti_32': 0.3247753252139955, 'lat_1080ti_64': 0.1805640315341493, 'lat_2080ti_1': 0.4278746215464485, 'lat_2080ti_32': 0.3205655131854198, 'lat_2080ti_64': 0.23270203262901976, 'lat_essential_ph_1': 0.2830188679245283, 'lat_eyeriss': 0.2106247593200448, 'lat_fpga': 0.24589596176490958, 'lat_gold_6226': 0.16790564735755714, 'lat_gold_6240': 0.32358707455696745, 'lat_pixel2': 0.21739130434782608, 'lat_pixel3': 0.21876113911960393, 'lat_raspi4': 0.26214602954248356, 'lat_samsung_a50': 0.09473684210526316, 'lat_samsung_s7': 0.14173228346456693, 'lat_silver_4114': 0.3322694540185846, 'lat_silver_4210r': 0.384015417606285, 'lat_titan_rtx_1': 0.4214638979419876, 'lat_titan_rtx_32': 0.33875989891371544, 'lat_titan_rtx_64': 0.24793987552573607, 'lat_titanx_1': 0.2311117891485737, 'lat_titanx_32': 0.2637854840951682, 'lat_titanx_64': 0.18463967717559188, 'lat_titanxp_1': 0.3957187062485045, 'lat_titanxp_32': 0.32572753429756895, 'lat_titanxp_64': 0.21468723434286274}
| |
FBNet_4594
|
FBNet
|
4594
|
4594
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_569[FLOAT, 16x3x3x3]
%onnx::Conv_570[FLOAT, 16]
%onnx::Conv_572[FLOAT, 96x16x1x1]
%onnx::Conv_573[FLOAT, 96]
%onnx::Conv_575[FLOAT, 96x1x5x5]
%onnx::Conv_578[FLOAT, 16x96x1x1]
%onnx::Conv_581[FLOAT, 24x16x1x1]
%onnx::Conv_582[FLOAT, 24]
%onnx::Conv_584[FLOAT, 144x24x1x1]
%onnx::Conv_585[FLOAT, 144]
%onnx::Conv_587[FLOAT, 144x1x5x5]
%onnx::Conv_590[FLOAT, 24x144x1x1]
%onnx::Conv_593[FLOAT, 24x24x1x1]
%onnx::Conv_596[FLOAT, 24x1x5x5]
%onnx::Conv_599[FLOAT, 24x24x1x1]
%onnx::Conv_602[FLOAT, 144x24x1x1]
%onnx::Conv_605[FLOAT, 144x1x3x3]
%onnx::Conv_608[FLOAT, 32x144x1x1]
%onnx::Conv_609[FLOAT, 32]
%onnx::Conv_611[FLOAT, 32x32x1x1]
%onnx::Conv_614[FLOAT, 32x1x3x3]
%onnx::Conv_617[FLOAT, 32x32x1x1]
%onnx::Conv_620[FLOAT, 192x32x1x1]
%onnx::Conv_621[FLOAT, 192]
%onnx::Conv_623[FLOAT, 192x1x5x5]
%onnx::Conv_626[FLOAT, 32x192x1x1]
%onnx::Conv_629[FLOAT, 96x32x1x1]
%onnx::Conv_632[FLOAT, 96x1x3x3]
%onnx::Conv_635[FLOAT, 32x96x1x1]
%onnx::Conv_638[FLOAT, 32x16x1x1]
%onnx::Conv_641[FLOAT, 32x1x3x3]
%onnx::Conv_644[FLOAT, 64x16x1x1]
%onnx::Conv_645[FLOAT, 64]
%onnx::Conv_647[FLOAT, 384x64x1x1]
%onnx::Conv_648[FLOAT, 384]
%onnx::Conv_650[FLOAT, 384x1x3x3]
%onnx::Conv_653[FLOAT, 64x384x1x1]
%onnx::Conv_656[FLOAT, 384x64x1x1]
%onnx::Conv_659[FLOAT, 384x1x5x5]
%onnx::Conv_662[FLOAT, 64x384x1x1]
%onnx::Conv_665[FLOAT, 384x64x1x1]
%onnx::Conv_668[FLOAT, 384x1x5x5]
%onnx::Conv_671[FLOAT, 64x384x1x1]
%onnx::Conv_674[FLOAT, 64x64x1x1]
%onnx::Conv_677[FLOAT, 64x1x3x3]
%onnx::Conv_680[FLOAT, 112x64x1x1]
%onnx::Conv_681[FLOAT, 112]
%onnx::Conv_683[FLOAT, 112x112x1x1]
%onnx::Conv_686[FLOAT, 112x1x3x3]
%onnx::Conv_689[FLOAT, 112x112x1x1]
%onnx::Conv_692[FLOAT, 672x112x1x1]
%onnx::Conv_693[FLOAT, 672]
%onnx::Conv_695[FLOAT, 672x1x3x3]
%onnx::Conv_698[FLOAT, 112x672x1x1]
%onnx::Conv_701[FLOAT, 184x112x1x1]
%onnx::Conv_702[FLOAT, 184]
%onnx::Conv_704[FLOAT, 184x184x1x1]
%onnx::Conv_707[FLOAT, 184x1x5x5]
%onnx::Conv_710[FLOAT, 184x184x1x1]
%onnx::Conv_713[FLOAT, 1104x184x1x1]
%onnx::Conv_714[FLOAT, 1104]
%onnx::Conv_716[FLOAT, 1104x1x3x3]
%onnx::Conv_719[FLOAT, 184x1104x1x1]
%onnx::Conv_722[FLOAT, 184x92x1x1]
%onnx::Conv_725[FLOAT, 184x1x3x3]
%onnx::Conv_728[FLOAT, 184x92x1x1]
%onnx::Conv_731[FLOAT, 184x184x1x1]
%onnx::Conv_734[FLOAT, 184x1x5x5]
%onnx::Conv_737[FLOAT, 352x184x1x1]
%onnx::Conv_738[FLOAT, 352]
%onnx::Conv_740[FLOAT, 1504x352x1x1]
%onnx::Conv_741[FLOAT, 1504]
) {
%onnx::Conv_735 = Identity(%onnx::Conv_702)
%onnx::Conv_732 = Identity(%onnx::Conv_702)
%onnx::Conv_729 = Identity(%onnx::Conv_702)
%onnx::Conv_726 = Identity(%onnx::Conv_702)
%onnx::Conv_723 = Identity(%onnx::Conv_702)
%onnx::Conv_720 = Identity(%onnx::Conv_702)
%onnx::Conv_717 = Identity(%onnx::Conv_714)
%onnx::Conv_711 = Identity(%onnx::Conv_702)
%onnx::Conv_708 = Identity(%onnx::Conv_702)
%onnx::Conv_705 = Identity(%onnx::Conv_702)
%onnx::Conv_699 = Identity(%onnx::Conv_681)
%onnx::Conv_696 = Identity(%onnx::Conv_693)
%onnx::Conv_690 = Identity(%onnx::Conv_681)
%onnx::Conv_687 = Identity(%onnx::Conv_681)
%onnx::Conv_684 = Identity(%onnx::Conv_681)
%onnx::Conv_678 = Identity(%onnx::Conv_645)
%onnx::Conv_675 = Identity(%onnx::Conv_645)
%onnx::Conv_672 = Identity(%onnx::Conv_645)
%onnx::Conv_669 = Identity(%onnx::Conv_648)
%onnx::Conv_666 = Identity(%onnx::Conv_648)
%onnx::Conv_663 = Identity(%onnx::Conv_645)
%onnx::Conv_660 = Identity(%onnx::Conv_648)
%onnx::Conv_657 = Identity(%onnx::Conv_648)
%onnx::Conv_654 = Identity(%onnx::Conv_645)
%onnx::Conv_651 = Identity(%onnx::Conv_648)
%onnx::Conv_642 = Identity(%onnx::Conv_609)
%onnx::Conv_639 = Identity(%onnx::Conv_609)
%onnx::Conv_636 = Identity(%onnx::Conv_609)
%onnx::Conv_633 = Identity(%onnx::Conv_573)
%onnx::Conv_630 = Identity(%onnx::Conv_573)
%onnx::Conv_627 = Identity(%onnx::Conv_609)
%onnx::Conv_624 = Identity(%onnx::Conv_621)
%onnx::Conv_618 = Identity(%onnx::Conv_609)
%onnx::Conv_615 = Identity(%onnx::Conv_609)
%onnx::Conv_612 = Identity(%onnx::Conv_609)
%onnx::Conv_606 = Identity(%onnx::Conv_585)
%onnx::Conv_603 = Identity(%onnx::Conv_585)
%onnx::Conv_600 = Identity(%onnx::Conv_582)
%onnx::Conv_597 = Identity(%onnx::Conv_582)
%onnx::Conv_594 = Identity(%onnx::Conv_582)
%onnx::Conv_591 = Identity(%onnx::Conv_582)
%onnx::Conv_588 = Identity(%onnx::Conv_585)
%onnx::Conv_579 = Identity(%onnx::Conv_570)
%onnx::Conv_576 = Identity(%onnx::Conv_573)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_569, %onnx::Conv_570)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_572, %onnx::Conv_573)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_575, %onnx::Conv_576)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_578, %onnx::Conv_579)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_581, %onnx::Conv_582)
%/cells.1/relu/Relu_output_0 = Relu(%/cells.1/conv/Conv_output_0)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/relu/Relu_output_0, %onnx::Conv_584, %onnx::Conv_585)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_587, %onnx::Conv_588)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_590, %onnx::Conv_591)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/relu/Relu_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_593, %onnx::Conv_594)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_596, %onnx::Conv_597)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_599, %onnx::Conv_600)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_602, %onnx::Conv_603)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_605, %onnx::Conv_606)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_608, %onnx::Conv_609)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_611, %onnx::Conv_612)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_614, %onnx::Conv_615)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_617, %onnx::Conv_618)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_620, %onnx::Conv_621)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_623, %onnx::Conv_624)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%567 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %567
}
|
val_accuracy
| 0
| 76,843,008
| 1,773,060
|
{'zcp_synflow': 76.02933768082389, 'zcp_zen': 66.06100463867188, 'zcp_epe_nas': 21.4932317807391, 'zcp_fisher': 0.17587727308273315, 'zcp_flops': 76843008.0, 'zcp_grad_norm': 25.15186882019043, 'zcp_grasp': 0.042163848876953125, 'zcp_jacov': -16.057210853296972, 'zcp_l2_norm': 623.8532104492188, 'zcp_nwot': 215.53833055723953, 'zcp_params': 1773060.0, 'zcp_plain': 0.00403332756832242, 'zcp_snip': 43.85196304321289, 'lat_1080ti_1': 0.43061336592762917, 'lat_1080ti_32': 0.5243345715757732, 'lat_1080ti_64': 0.5387760041524292, 'lat_2080ti_1': 0.5085566446801452, 'lat_2080ti_32': 0.4911309768379632, 'lat_2080ti_64': 0.5336540397528611, 'lat_essential_ph_1': 0.33962264150943394, 'lat_eyeriss': 0.5799515889461428, 'lat_fpga': 0.5379134960765493, 'lat_gold_6226': 0.3842700090330471, 'lat_gold_6240': 0.4355334316941717, 'lat_pixel2': 0.34782608695652173, 'lat_pixel3': 0.5681417552433963, 'lat_raspi4': 0.5430320922765366, 'lat_samsung_a50': 0.21052631578947367, 'lat_samsung_s7': 0.18110236220472442, 'lat_silver_4114': 0.5051536669740312, 'lat_silver_4210r': 0.4358880142605662, 'lat_titan_rtx_1': 0.4189153770072189, 'lat_titan_rtx_32': 0.4461825610033891, 'lat_titan_rtx_64': 0.5167555483050909, 'lat_titanx_1': 0.23194074052980174, 'lat_titanx_32': 0.4850541617878564, 'lat_titanx_64': 0.5078085448279734, 'lat_titanxp_1': 0.3926860738522562, 'lat_titanxp_32': 0.4887621114933278, 'lat_titanxp_64': 0.5292402717440946}
| |
FBNet_644
|
FBNet
|
644
|
644
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_651[FLOAT, 16x3x3x3]
%onnx::Conv_652[FLOAT, 16]
%onnx::Conv_654[FLOAT, 16x8x1x1]
%onnx::Conv_657[FLOAT, 16x1x3x3]
%onnx::Conv_660[FLOAT, 16x8x1x1]
%onnx::Conv_663[FLOAT, 24x16x1x1]
%onnx::Conv_664[FLOAT, 24]
%onnx::Conv_666[FLOAT, 24x12x1x1]
%onnx::Conv_669[FLOAT, 24x1x5x5]
%onnx::Conv_672[FLOAT, 24x12x1x1]
%onnx::Conv_675[FLOAT, 24x24x1x1]
%onnx::Conv_678[FLOAT, 24x1x5x5]
%onnx::Conv_681[FLOAT, 24x24x1x1]
%onnx::Conv_684[FLOAT, 24x24x1x1]
%onnx::Conv_687[FLOAT, 24x1x5x5]
%onnx::Conv_690[FLOAT, 24x24x1x1]
%onnx::Conv_693[FLOAT, 24x24x1x1]
%onnx::Conv_696[FLOAT, 24x1x3x3]
%onnx::Conv_699[FLOAT, 32x24x1x1]
%onnx::Conv_700[FLOAT, 32]
%onnx::Conv_702[FLOAT, 192x32x1x1]
%onnx::Conv_703[FLOAT, 192]
%onnx::Conv_705[FLOAT, 192x1x5x5]
%onnx::Conv_708[FLOAT, 32x192x1x1]
%onnx::Conv_711[FLOAT, 192x32x1x1]
%onnx::Conv_714[FLOAT, 192x1x3x3]
%onnx::Conv_717[FLOAT, 32x192x1x1]
%onnx::Conv_720[FLOAT, 192x32x1x1]
%onnx::Conv_723[FLOAT, 192x1x5x5]
%onnx::Conv_726[FLOAT, 32x192x1x1]
%onnx::Conv_729[FLOAT, 96x32x1x1]
%onnx::Conv_730[FLOAT, 96]
%onnx::Conv_732[FLOAT, 96x1x3x3]
%onnx::Conv_735[FLOAT, 64x96x1x1]
%onnx::Conv_736[FLOAT, 64]
%onnx::Conv_738[FLOAT, 64x32x1x1]
%onnx::Conv_741[FLOAT, 64x1x5x5]
%onnx::Conv_744[FLOAT, 64x32x1x1]
%onnx::Conv_747[FLOAT, 192x64x1x1]
%onnx::Conv_750[FLOAT, 192x1x5x5]
%onnx::Conv_753[FLOAT, 64x192x1x1]
%onnx::Conv_756[FLOAT, 64x64x1x1]
%onnx::Conv_759[FLOAT, 64x1x3x3]
%onnx::Conv_762[FLOAT, 112x64x1x1]
%onnx::Conv_763[FLOAT, 112]
%onnx::Conv_765[FLOAT, 336x112x1x1]
%onnx::Conv_766[FLOAT, 336]
%onnx::Conv_768[FLOAT, 336x1x3x3]
%onnx::Conv_771[FLOAT, 112x336x1x1]
%onnx::Conv_774[FLOAT, 112x112x1x1]
%onnx::Conv_777[FLOAT, 112x1x5x5]
%onnx::Conv_780[FLOAT, 112x112x1x1]
%onnx::Conv_783[FLOAT, 112x112x1x1]
%onnx::Conv_786[FLOAT, 112x1x5x5]
%onnx::Conv_789[FLOAT, 112x112x1x1]
%onnx::Conv_792[FLOAT, 112x56x1x1]
%onnx::Conv_795[FLOAT, 112x1x5x5]
%onnx::Conv_798[FLOAT, 184x56x1x1]
%onnx::Conv_799[FLOAT, 184]
%onnx::Conv_801[FLOAT, 1104x184x1x1]
%onnx::Conv_802[FLOAT, 1104]
%onnx::Conv_804[FLOAT, 1104x1x5x5]
%onnx::Conv_807[FLOAT, 184x1104x1x1]
%onnx::Conv_810[FLOAT, 184x184x1x1]
%onnx::Conv_813[FLOAT, 184x1x5x5]
%onnx::Conv_816[FLOAT, 184x184x1x1]
%onnx::Conv_819[FLOAT, 184x184x1x1]
%onnx::Conv_822[FLOAT, 184x1x5x5]
%onnx::Conv_825[FLOAT, 184x184x1x1]
%onnx::Conv_828[FLOAT, 1104x184x1x1]
%onnx::Conv_831[FLOAT, 1104x1x3x3]
%onnx::Conv_834[FLOAT, 352x1104x1x1]
%onnx::Conv_835[FLOAT, 352]
%onnx::Conv_837[FLOAT, 1504x352x1x1]
%onnx::Conv_838[FLOAT, 1504]
) {
%onnx::Conv_832 = Identity(%onnx::Conv_802)
%onnx::Conv_829 = Identity(%onnx::Conv_802)
%onnx::Conv_826 = Identity(%onnx::Conv_799)
%onnx::Conv_823 = Identity(%onnx::Conv_799)
%onnx::Conv_820 = Identity(%onnx::Conv_799)
%onnx::Conv_817 = Identity(%onnx::Conv_799)
%onnx::Conv_814 = Identity(%onnx::Conv_799)
%onnx::Conv_811 = Identity(%onnx::Conv_799)
%onnx::Conv_808 = Identity(%onnx::Conv_799)
%onnx::Conv_805 = Identity(%onnx::Conv_802)
%onnx::Conv_796 = Identity(%onnx::Conv_763)
%onnx::Conv_793 = Identity(%onnx::Conv_763)
%onnx::Conv_790 = Identity(%onnx::Conv_763)
%onnx::Conv_787 = Identity(%onnx::Conv_763)
%onnx::Conv_784 = Identity(%onnx::Conv_763)
%onnx::Conv_781 = Identity(%onnx::Conv_763)
%onnx::Conv_778 = Identity(%onnx::Conv_763)
%onnx::Conv_775 = Identity(%onnx::Conv_763)
%onnx::Conv_772 = Identity(%onnx::Conv_763)
%onnx::Conv_769 = Identity(%onnx::Conv_766)
%onnx::Conv_760 = Identity(%onnx::Conv_736)
%onnx::Conv_757 = Identity(%onnx::Conv_736)
%onnx::Conv_754 = Identity(%onnx::Conv_736)
%onnx::Conv_751 = Identity(%onnx::Conv_703)
%onnx::Conv_748 = Identity(%onnx::Conv_703)
%onnx::Conv_745 = Identity(%onnx::Conv_736)
%onnx::Conv_742 = Identity(%onnx::Conv_736)
%onnx::Conv_739 = Identity(%onnx::Conv_736)
%onnx::Conv_733 = Identity(%onnx::Conv_730)
%onnx::Conv_727 = Identity(%onnx::Conv_700)
%onnx::Conv_724 = Identity(%onnx::Conv_703)
%onnx::Conv_721 = Identity(%onnx::Conv_703)
%onnx::Conv_718 = Identity(%onnx::Conv_700)
%onnx::Conv_715 = Identity(%onnx::Conv_703)
%onnx::Conv_712 = Identity(%onnx::Conv_703)
%onnx::Conv_709 = Identity(%onnx::Conv_700)
%onnx::Conv_706 = Identity(%onnx::Conv_703)
%onnx::Conv_697 = Identity(%onnx::Conv_664)
%onnx::Conv_694 = Identity(%onnx::Conv_664)
%onnx::Conv_691 = Identity(%onnx::Conv_664)
%onnx::Conv_688 = Identity(%onnx::Conv_664)
%onnx::Conv_685 = Identity(%onnx::Conv_664)
%onnx::Conv_682 = Identity(%onnx::Conv_664)
%onnx::Conv_679 = Identity(%onnx::Conv_664)
%onnx::Conv_676 = Identity(%onnx::Conv_664)
%onnx::Conv_673 = Identity(%onnx::Conv_664)
%onnx::Conv_670 = Identity(%onnx::Conv_664)
%onnx::Conv_667 = Identity(%onnx::Conv_664)
%onnx::Conv_661 = Identity(%onnx::Conv_652)
%onnx::Conv_658 = Identity(%onnx::Conv_652)
%onnx::Conv_655 = Identity(%onnx::Conv_652)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_651, %onnx::Conv_652)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.1/relu/Relu_output_0 = Relu(%/cells.1/conv/Conv_output_0)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/relu/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/relu/Relu_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%649 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %649
}
|
val_accuracy
| 0
| 62,560,640
| 2,150,548
|
{'zcp_synflow': 79.93784910567355, 'zcp_zen': 69.06388092041016, 'zcp_epe_nas': 18.165278711090867, 'zcp_fisher': 0.11444547027349472, 'zcp_flops': 62560640.0, 'zcp_grad_norm': 22.748554229736328, 'zcp_grasp': 0.07481575012207031, 'zcp_jacov': -16.054634652045507, 'zcp_l2_norm': 626.951904296875, 'zcp_nwot': 208.18874947668368, 'zcp_params': 2150548.0, 'zcp_plain': 0.0023520332761108875, 'zcp_snip': 34.6345100402832, 'lat_1080ti_1': 0.650992903354702, 'lat_1080ti_32': 0.4655540908822625, 'lat_1080ti_64': 0.3017071060003364, 'lat_2080ti_1': 0.6135494823926089, 'lat_2080ti_32': 0.4589310528544911, 'lat_2080ti_64': 0.3357890445700378, 'lat_essential_ph_1': 0.3018867924528302, 'lat_eyeriss': 0.3977408174866593, 'lat_fpga': 0.41145271574031006, 'lat_gold_6226': 0.3244590413202777, 'lat_gold_6240': 0.6260352062707764, 'lat_pixel2': 0.2391304347826087, 'lat_pixel3': 0.3874792218606622, 'lat_raspi4': 0.4582854439056419, 'lat_samsung_a50': 0.16842105263157894, 'lat_samsung_s7': 0.15748031496062992, 'lat_silver_4114': 0.6465451389568225, 'lat_silver_4210r': 0.6489644489750035, 'lat_titan_rtx_1': 0.5904695216316534, 'lat_titan_rtx_32': 0.4933531149901347, 'lat_titan_rtx_64': 0.3626176989597341, 'lat_titanx_1': 0.32522055269630024, 'lat_titanx_32': 0.38230216641292003, 'lat_titanx_64': 0.3087067887920109, 'lat_titanxp_1': 0.5773260646286174, 'lat_titanxp_32': 0.43545276595421517, 'lat_titanxp_64': 0.33317109597685957}
| |
FBNet_2364
|
FBNet
|
2364
|
2364
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_641[FLOAT, 16x3x3x3]
%onnx::Conv_642[FLOAT, 16]
%onnx::Conv_644[FLOAT, 16x8x1x1]
%onnx::Conv_647[FLOAT, 16x1x5x5]
%onnx::Conv_650[FLOAT, 16x8x1x1]
%onnx::Conv_653[FLOAT, 96x16x1x1]
%onnx::Conv_654[FLOAT, 96]
%onnx::Conv_656[FLOAT, 96x1x3x3]
%onnx::Conv_659[FLOAT, 24x96x1x1]
%onnx::Conv_660[FLOAT, 24]
%onnx::Conv_662[FLOAT, 24x24x1x1]
%onnx::Conv_665[FLOAT, 24x1x5x5]
%onnx::Conv_668[FLOAT, 24x24x1x1]
%onnx::Conv_671[FLOAT, 72x24x1x1]
%onnx::Conv_672[FLOAT, 72]
%onnx::Conv_674[FLOAT, 72x1x3x3]
%onnx::Conv_677[FLOAT, 24x72x1x1]
%onnx::Conv_680[FLOAT, 24x24x1x1]
%onnx::Conv_683[FLOAT, 24x1x3x3]
%onnx::Conv_686[FLOAT, 24x24x1x1]
%onnx::Conv_689[FLOAT, 72x24x1x1]
%onnx::Conv_692[FLOAT, 72x1x5x5]
%onnx::Conv_695[FLOAT, 32x72x1x1]
%onnx::Conv_696[FLOAT, 32]
%onnx::Conv_698[FLOAT, 32x32x1x1]
%onnx::Conv_701[FLOAT, 32x1x5x5]
%onnx::Conv_704[FLOAT, 32x32x1x1]
%onnx::Conv_707[FLOAT, 96x32x1x1]
%onnx::Conv_710[FLOAT, 96x1x3x3]
%onnx::Conv_713[FLOAT, 32x96x1x1]
%onnx::Conv_716[FLOAT, 192x32x1x1]
%onnx::Conv_717[FLOAT, 192]
%onnx::Conv_719[FLOAT, 192x1x5x5]
%onnx::Conv_722[FLOAT, 64x192x1x1]
%onnx::Conv_723[FLOAT, 64]
%onnx::Conv_725[FLOAT, 64x64x1x1]
%onnx::Conv_728[FLOAT, 64x1x3x3]
%onnx::Conv_731[FLOAT, 64x64x1x1]
%onnx::Conv_734[FLOAT, 192x64x1x1]
%onnx::Conv_737[FLOAT, 192x1x3x3]
%onnx::Conv_740[FLOAT, 64x192x1x1]
%onnx::Conv_743[FLOAT, 64x32x1x1]
%onnx::Conv_746[FLOAT, 64x1x5x5]
%onnx::Conv_749[FLOAT, 64x32x1x1]
%onnx::Conv_752[FLOAT, 64x32x1x1]
%onnx::Conv_755[FLOAT, 64x1x3x3]
%onnx::Conv_758[FLOAT, 112x32x1x1]
%onnx::Conv_759[FLOAT, 112]
%onnx::Conv_761[FLOAT, 336x112x1x1]
%onnx::Conv_762[FLOAT, 336]
%onnx::Conv_764[FLOAT, 336x1x5x5]
%onnx::Conv_767[FLOAT, 112x336x1x1]
%onnx::Conv_770[FLOAT, 336x112x1x1]
%onnx::Conv_773[FLOAT, 336x1x3x3]
%onnx::Conv_776[FLOAT, 112x336x1x1]
%onnx::Conv_779[FLOAT, 184x112x1x1]
%onnx::Conv_780[FLOAT, 184]
%onnx::Conv_782[FLOAT, 184x92x1x1]
%onnx::Conv_785[FLOAT, 184x1x5x5]
%onnx::Conv_788[FLOAT, 184x92x1x1]
%onnx::Conv_791[FLOAT, 1104x184x1x1]
%onnx::Conv_792[FLOAT, 1104]
%onnx::Conv_794[FLOAT, 1104x1x3x3]
%onnx::Conv_797[FLOAT, 184x1104x1x1]
%onnx::Conv_800[FLOAT, 1104x184x1x1]
%onnx::Conv_803[FLOAT, 1104x1x3x3]
%onnx::Conv_806[FLOAT, 184x1104x1x1]
%onnx::Conv_809[FLOAT, 184x92x1x1]
%onnx::Conv_812[FLOAT, 184x1x5x5]
%onnx::Conv_815[FLOAT, 352x92x1x1]
%onnx::Conv_816[FLOAT, 352]
%onnx::Conv_818[FLOAT, 1504x352x1x1]
%onnx::Conv_819[FLOAT, 1504]
) {
%onnx::Conv_813 = Identity(%onnx::Conv_780)
%onnx::Conv_810 = Identity(%onnx::Conv_780)
%onnx::Conv_807 = Identity(%onnx::Conv_780)
%onnx::Conv_804 = Identity(%onnx::Conv_792)
%onnx::Conv_801 = Identity(%onnx::Conv_792)
%onnx::Conv_798 = Identity(%onnx::Conv_780)
%onnx::Conv_795 = Identity(%onnx::Conv_792)
%onnx::Conv_789 = Identity(%onnx::Conv_780)
%onnx::Conv_786 = Identity(%onnx::Conv_780)
%onnx::Conv_783 = Identity(%onnx::Conv_780)
%onnx::Conv_777 = Identity(%onnx::Conv_759)
%onnx::Conv_774 = Identity(%onnx::Conv_762)
%onnx::Conv_771 = Identity(%onnx::Conv_762)
%onnx::Conv_768 = Identity(%onnx::Conv_759)
%onnx::Conv_765 = Identity(%onnx::Conv_762)
%onnx::Conv_756 = Identity(%onnx::Conv_723)
%onnx::Conv_753 = Identity(%onnx::Conv_723)
%onnx::Conv_750 = Identity(%onnx::Conv_723)
%onnx::Conv_747 = Identity(%onnx::Conv_723)
%onnx::Conv_744 = Identity(%onnx::Conv_723)
%onnx::Conv_741 = Identity(%onnx::Conv_723)
%onnx::Conv_738 = Identity(%onnx::Conv_717)
%onnx::Conv_735 = Identity(%onnx::Conv_717)
%onnx::Conv_732 = Identity(%onnx::Conv_723)
%onnx::Conv_729 = Identity(%onnx::Conv_723)
%onnx::Conv_726 = Identity(%onnx::Conv_723)
%onnx::Conv_720 = Identity(%onnx::Conv_717)
%onnx::Conv_714 = Identity(%onnx::Conv_696)
%onnx::Conv_711 = Identity(%onnx::Conv_654)
%onnx::Conv_708 = Identity(%onnx::Conv_654)
%onnx::Conv_705 = Identity(%onnx::Conv_696)
%onnx::Conv_702 = Identity(%onnx::Conv_696)
%onnx::Conv_699 = Identity(%onnx::Conv_696)
%onnx::Conv_693 = Identity(%onnx::Conv_672)
%onnx::Conv_690 = Identity(%onnx::Conv_672)
%onnx::Conv_687 = Identity(%onnx::Conv_660)
%onnx::Conv_684 = Identity(%onnx::Conv_660)
%onnx::Conv_681 = Identity(%onnx::Conv_660)
%onnx::Conv_678 = Identity(%onnx::Conv_660)
%onnx::Conv_675 = Identity(%onnx::Conv_672)
%onnx::Conv_669 = Identity(%onnx::Conv_660)
%onnx::Conv_666 = Identity(%onnx::Conv_660)
%onnx::Conv_663 = Identity(%onnx::Conv_660)
%onnx::Conv_657 = Identity(%onnx::Conv_654)
%onnx::Conv_651 = Identity(%onnx::Conv_642)
%onnx::Conv_648 = Identity(%onnx::Conv_642)
%onnx::Conv_645 = Identity(%onnx::Conv_642)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_641, %onnx::Conv_642)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_output_0 = Reshape(%/cells.18/nl/Relu_output_0, %/cells.18/shuffle/Constant_output_0)
%/cells.18/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.18/shuffle/Reshape_output_0)
%/cells.18/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_1_output_0 = Reshape(%/cells.18/shuffle/Transpose_output_0, %/cells.18/shuffle/Constant_1_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/shuffle/Reshape_1_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_803, %onnx::Conv_804)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_806, %onnx::Conv_807)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_809, %onnx::Conv_810)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_812, %onnx::Conv_813)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_815, %onnx::Conv_816)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_818, %onnx::Conv_819)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%639 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %639
}
|
val_accuracy
| 0
| 61,227,776
| 1,910,004
|
{'zcp_synflow': 73.77133543218883, 'zcp_zen': 63.85462951660156, 'zcp_epe_nas': 13.603699385412378, 'zcp_fisher': 0.08176301419734955, 'zcp_flops': 61227776.0, 'zcp_grad_norm': 21.835596084594727, 'zcp_grasp': -0.057712554931640625, 'zcp_jacov': -16.04914851448025, 'zcp_l2_norm': 585.2902221679688, 'zcp_nwot': 210.66323439894794, 'zcp_params': 1910004.0, 'zcp_plain': -0.004843680653721094, 'zcp_snip': 39.855159759521484, 'lat_1080ti_1': 0.4919300093558454, 'lat_1080ti_32': 0.42858637639731234, 'lat_1080ti_64': 0.32469763277393493, 'lat_2080ti_1': 0.5555973273412533, 'lat_2080ti_32': 0.4593551114351828, 'lat_2080ti_64': 0.35182237249811155, 'lat_essential_ph_1': 0.39622641509433965, 'lat_eyeriss': 0.35095263418480566, 'lat_fpga': 0.3977181646364994, 'lat_gold_6226': 0.3364634835832208, 'lat_gold_6240': 0.48789985049674556, 'lat_pixel2': 0.391304347826087, 'lat_pixel3': 0.31977012061238563, 'lat_raspi4': 0.3744446352349925, 'lat_samsung_a50': 0.17894736842105263, 'lat_samsung_s7': 0.16535433070866143, 'lat_silver_4114': 0.5089091510812164, 'lat_silver_4210r': 0.51038713202127, 'lat_titan_rtx_1': 0.5324519476123962, 'lat_titan_rtx_32': 0.4591534823221122, 'lat_titan_rtx_64': 0.35881394999692395, 'lat_titanx_1': 0.28719832089020103, 'lat_titanx_32': 0.3896486955149152, 'lat_titanx_64': 0.31711869548179034, 'lat_titanxp_1': 0.5023957045721924, 'lat_titanxp_32': 0.4274154969602258, 'lat_titanxp_64': 0.34946331229465033}
| |
FBNet_767
|
FBNet
|
767
|
767
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_697[FLOAT, 16x3x3x3]
%onnx::Conv_698[FLOAT, 16]
%onnx::Conv_700[FLOAT, 16x8x1x1]
%onnx::Conv_703[FLOAT, 16x1x5x5]
%onnx::Conv_706[FLOAT, 16x8x1x1]
%onnx::Conv_709[FLOAT, 16x16x1x1]
%onnx::Conv_712[FLOAT, 16x1x5x5]
%onnx::Conv_715[FLOAT, 24x16x1x1]
%onnx::Conv_716[FLOAT, 24]
%onnx::Conv_718[FLOAT, 72x24x1x1]
%onnx::Conv_719[FLOAT, 72]
%onnx::Conv_721[FLOAT, 72x1x3x3]
%onnx::Conv_724[FLOAT, 24x72x1x1]
%onnx::Conv_727[FLOAT, 24x12x1x1]
%onnx::Conv_730[FLOAT, 24x1x3x3]
%onnx::Conv_733[FLOAT, 24x12x1x1]
%onnx::Conv_736[FLOAT, 24x12x1x1]
%onnx::Conv_739[FLOAT, 24x1x5x5]
%onnx::Conv_742[FLOAT, 24x12x1x1]
%onnx::Conv_745[FLOAT, 72x24x1x1]
%onnx::Conv_748[FLOAT, 72x1x3x3]
%onnx::Conv_751[FLOAT, 32x72x1x1]
%onnx::Conv_752[FLOAT, 32]
%onnx::Conv_754[FLOAT, 32x32x1x1]
%onnx::Conv_757[FLOAT, 32x1x5x5]
%onnx::Conv_760[FLOAT, 32x32x1x1]
%onnx::Conv_763[FLOAT, 32x16x1x1]
%onnx::Conv_766[FLOAT, 32x1x3x3]
%onnx::Conv_769[FLOAT, 32x16x1x1]
%onnx::Conv_772[FLOAT, 32x16x1x1]
%onnx::Conv_775[FLOAT, 32x1x3x3]
%onnx::Conv_778[FLOAT, 32x16x1x1]
%onnx::Conv_781[FLOAT, 64x32x1x1]
%onnx::Conv_782[FLOAT, 64]
%onnx::Conv_784[FLOAT, 64x32x1x1]
%onnx::Conv_787[FLOAT, 64x1x5x5]
%onnx::Conv_790[FLOAT, 64x32x1x1]
%onnx::Conv_793[FLOAT, 384x64x1x1]
%onnx::Conv_794[FLOAT, 384]
%onnx::Conv_796[FLOAT, 384x1x3x3]
%onnx::Conv_799[FLOAT, 64x384x1x1]
%onnx::Conv_802[FLOAT, 64x32x1x1]
%onnx::Conv_805[FLOAT, 64x1x5x5]
%onnx::Conv_808[FLOAT, 64x32x1x1]
%onnx::Conv_811[FLOAT, 384x64x1x1]
%onnx::Conv_814[FLOAT, 384x1x3x3]
%onnx::Conv_817[FLOAT, 112x384x1x1]
%onnx::Conv_818[FLOAT, 112]
%onnx::Conv_820[FLOAT, 112x56x1x1]
%onnx::Conv_823[FLOAT, 112x1x5x5]
%onnx::Conv_826[FLOAT, 112x56x1x1]
%onnx::Conv_829[FLOAT, 336x112x1x1]
%onnx::Conv_830[FLOAT, 336]
%onnx::Conv_832[FLOAT, 336x1x3x3]
%onnx::Conv_835[FLOAT, 112x336x1x1]
%onnx::Conv_838[FLOAT, 112x112x1x1]
%onnx::Conv_841[FLOAT, 112x1x3x3]
%onnx::Conv_844[FLOAT, 112x112x1x1]
%onnx::Conv_847[FLOAT, 672x112x1x1]
%onnx::Conv_848[FLOAT, 672]
%onnx::Conv_850[FLOAT, 672x1x5x5]
%onnx::Conv_853[FLOAT, 184x672x1x1]
%onnx::Conv_854[FLOAT, 184]
%onnx::Conv_856[FLOAT, 184x184x1x1]
%onnx::Conv_859[FLOAT, 184x1x3x3]
%onnx::Conv_862[FLOAT, 184x184x1x1]
%onnx::Conv_865[FLOAT, 552x184x1x1]
%onnx::Conv_866[FLOAT, 552]
%onnx::Conv_868[FLOAT, 552x1x5x5]
%onnx::Conv_871[FLOAT, 352x552x1x1]
%onnx::Conv_872[FLOAT, 352]
%onnx::Conv_874[FLOAT, 1504x352x1x1]
%onnx::Conv_875[FLOAT, 1504]
) {
%onnx::Conv_869 = Identity(%onnx::Conv_866)
%onnx::Conv_863 = Identity(%onnx::Conv_854)
%onnx::Conv_860 = Identity(%onnx::Conv_854)
%onnx::Conv_857 = Identity(%onnx::Conv_854)
%onnx::Conv_851 = Identity(%onnx::Conv_848)
%onnx::Conv_845 = Identity(%onnx::Conv_818)
%onnx::Conv_842 = Identity(%onnx::Conv_818)
%onnx::Conv_839 = Identity(%onnx::Conv_818)
%onnx::Conv_836 = Identity(%onnx::Conv_818)
%onnx::Conv_833 = Identity(%onnx::Conv_830)
%onnx::Conv_827 = Identity(%onnx::Conv_818)
%onnx::Conv_824 = Identity(%onnx::Conv_818)
%onnx::Conv_821 = Identity(%onnx::Conv_818)
%onnx::Conv_815 = Identity(%onnx::Conv_794)
%onnx::Conv_812 = Identity(%onnx::Conv_794)
%onnx::Conv_809 = Identity(%onnx::Conv_782)
%onnx::Conv_806 = Identity(%onnx::Conv_782)
%onnx::Conv_803 = Identity(%onnx::Conv_782)
%onnx::Conv_800 = Identity(%onnx::Conv_782)
%onnx::Conv_797 = Identity(%onnx::Conv_794)
%onnx::Conv_791 = Identity(%onnx::Conv_782)
%onnx::Conv_788 = Identity(%onnx::Conv_782)
%onnx::Conv_785 = Identity(%onnx::Conv_782)
%onnx::Conv_779 = Identity(%onnx::Conv_752)
%onnx::Conv_776 = Identity(%onnx::Conv_752)
%onnx::Conv_773 = Identity(%onnx::Conv_752)
%onnx::Conv_770 = Identity(%onnx::Conv_752)
%onnx::Conv_767 = Identity(%onnx::Conv_752)
%onnx::Conv_764 = Identity(%onnx::Conv_752)
%onnx::Conv_761 = Identity(%onnx::Conv_752)
%onnx::Conv_758 = Identity(%onnx::Conv_752)
%onnx::Conv_755 = Identity(%onnx::Conv_752)
%onnx::Conv_749 = Identity(%onnx::Conv_719)
%onnx::Conv_746 = Identity(%onnx::Conv_719)
%onnx::Conv_743 = Identity(%onnx::Conv_716)
%onnx::Conv_740 = Identity(%onnx::Conv_716)
%onnx::Conv_737 = Identity(%onnx::Conv_716)
%onnx::Conv_734 = Identity(%onnx::Conv_716)
%onnx::Conv_731 = Identity(%onnx::Conv_716)
%onnx::Conv_728 = Identity(%onnx::Conv_716)
%onnx::Conv_725 = Identity(%onnx::Conv_716)
%onnx::Conv_722 = Identity(%onnx::Conv_719)
%onnx::Conv_713 = Identity(%onnx::Conv_698)
%onnx::Conv_710 = Identity(%onnx::Conv_698)
%onnx::Conv_707 = Identity(%onnx::Conv_698)
%onnx::Conv_704 = Identity(%onnx::Conv_698)
%onnx::Conv_701 = Identity(%onnx::Conv_698)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_697, %onnx::Conv_698)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.8/Add_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.9/relu/Relu_output_0 = Relu(%/cells.9/conv/Conv_output_0)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/relu/Relu_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/relu/Relu_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_823, %onnx::Conv_824)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_826, %onnx::Conv_827)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_829, %onnx::Conv_830)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_832, %onnx::Conv_833)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_835, %onnx::Conv_836)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_838, %onnx::Conv_839)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_841, %onnx::Conv_842)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_844, %onnx::Conv_845)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_847, %onnx::Conv_848)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_850, %onnx::Conv_851)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_853, %onnx::Conv_854)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_856, %onnx::Conv_857)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_859, %onnx::Conv_860)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_862, %onnx::Conv_863)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_865, %onnx::Conv_866)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_868, %onnx::Conv_869)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_871, %onnx::Conv_872)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_874, %onnx::Conv_875)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%695 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %695
}
|
val_accuracy
| 0
| 52,813,184
| 1,568,932
|
{'zcp_synflow': 69.7073667564458, 'zcp_zen': 59.78759002685547, 'zcp_epe_nas': 11.134283034275674, 'zcp_fisher': 0.07636605948209763, 'zcp_flops': 52813184.0, 'zcp_grad_norm': 20.542192459106445, 'zcp_grasp': -0.06719017028808594, 'zcp_jacov': -16.069074058471188, 'zcp_l2_norm': 524.7709350585938, 'zcp_nwot': 207.0891592328685, 'zcp_params': 1568932.0, 'zcp_plain': -0.0030286763794720173, 'zcp_snip': 30.951807022094727, 'lat_1080ti_1': 0.6041467290722288, 'lat_1080ti_32': 0.4402007015317937, 'lat_1080ti_64': 0.2721892973638171, 'lat_2080ti_1': 0.6011601787357238, 'lat_2080ti_32': 0.48840690944175735, 'lat_2080ti_64': 0.2963186020061607, 'lat_essential_ph_1': 0.3018867924528302, 'lat_eyeriss': 0.21531916454257044, 'lat_fpga': 0.2468557971086196, 'lat_gold_6226': 0.21077085588520544, 'lat_gold_6240': 0.36325726220862564, 'lat_pixel2': 0.17391304347826086, 'lat_pixel3': 0.2488815542250712, 'lat_raspi4': 0.31870763452775536, 'lat_samsung_a50': 0.11578947368421053, 'lat_samsung_s7': 0.15748031496062992, 'lat_silver_4114': 0.4294311500462007, 'lat_silver_4210r': 0.3742278373350155, 'lat_titan_rtx_1': 0.5773277038904463, 'lat_titan_rtx_32': 0.48411782827759003, 'lat_titan_rtx_64': 0.3264489029132759, 'lat_titanx_1': 0.31823888808958084, 'lat_titanx_32': 0.3706473725376138, 'lat_titanx_64': 0.26161763343926236, 'lat_titanxp_1': 0.5601125021290732, 'lat_titanxp_32': 0.42610949845805857, 'lat_titanxp_64': 0.2946489975841702}
| |
FBNet_1451
|
FBNet
|
1451
|
1451
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_661[FLOAT, 16x3x3x3]
%onnx::Conv_662[FLOAT, 16]
%onnx::Conv_664[FLOAT, 16x16x1x1]
%onnx::Conv_667[FLOAT, 16x1x3x3]
%onnx::Conv_670[FLOAT, 16x16x1x1]
%onnx::Conv_673[FLOAT, 24x16x1x1]
%onnx::Conv_674[FLOAT, 24]
%onnx::Conv_676[FLOAT, 144x24x1x1]
%onnx::Conv_677[FLOAT, 144]
%onnx::Conv_679[FLOAT, 144x1x5x5]
%onnx::Conv_682[FLOAT, 24x144x1x1]
%onnx::Conv_685[FLOAT, 24x12x1x1]
%onnx::Conv_688[FLOAT, 24x1x3x3]
%onnx::Conv_691[FLOAT, 24x12x1x1]
%onnx::Conv_694[FLOAT, 144x24x1x1]
%onnx::Conv_697[FLOAT, 144x1x3x3]
%onnx::Conv_700[FLOAT, 24x144x1x1]
%onnx::Conv_703[FLOAT, 24x12x1x1]
%onnx::Conv_706[FLOAT, 24x1x3x3]
%onnx::Conv_709[FLOAT, 32x12x1x1]
%onnx::Conv_710[FLOAT, 32]
%onnx::Conv_712[FLOAT, 192x32x1x1]
%onnx::Conv_713[FLOAT, 192]
%onnx::Conv_715[FLOAT, 192x1x3x3]
%onnx::Conv_718[FLOAT, 32x192x1x1]
%onnx::Conv_721[FLOAT, 96x32x1x1]
%onnx::Conv_722[FLOAT, 96]
%onnx::Conv_724[FLOAT, 96x1x5x5]
%onnx::Conv_727[FLOAT, 32x96x1x1]
%onnx::Conv_730[FLOAT, 32x16x1x1]
%onnx::Conv_733[FLOAT, 32x1x3x3]
%onnx::Conv_736[FLOAT, 32x16x1x1]
%onnx::Conv_739[FLOAT, 192x32x1x1]
%onnx::Conv_742[FLOAT, 192x1x5x5]
%onnx::Conv_745[FLOAT, 64x192x1x1]
%onnx::Conv_746[FLOAT, 64]
%onnx::Conv_748[FLOAT, 192x64x1x1]
%onnx::Conv_751[FLOAT, 192x1x3x3]
%onnx::Conv_754[FLOAT, 64x192x1x1]
%onnx::Conv_757[FLOAT, 384x64x1x1]
%onnx::Conv_758[FLOAT, 384]
%onnx::Conv_760[FLOAT, 384x1x5x5]
%onnx::Conv_763[FLOAT, 64x384x1x1]
%onnx::Conv_766[FLOAT, 384x64x1x1]
%onnx::Conv_769[FLOAT, 384x1x3x3]
%onnx::Conv_772[FLOAT, 64x384x1x1]
%onnx::Conv_775[FLOAT, 64x32x1x1]
%onnx::Conv_778[FLOAT, 64x1x3x3]
%onnx::Conv_781[FLOAT, 112x32x1x1]
%onnx::Conv_782[FLOAT, 112]
%onnx::Conv_784[FLOAT, 336x112x1x1]
%onnx::Conv_785[FLOAT, 336]
%onnx::Conv_787[FLOAT, 336x1x3x3]
%onnx::Conv_790[FLOAT, 112x336x1x1]
%onnx::Conv_793[FLOAT, 112x56x1x1]
%onnx::Conv_796[FLOAT, 112x1x5x5]
%onnx::Conv_799[FLOAT, 184x56x1x1]
%onnx::Conv_800[FLOAT, 184]
%onnx::Conv_802[FLOAT, 184x184x1x1]
%onnx::Conv_805[FLOAT, 184x1x5x5]
%onnx::Conv_808[FLOAT, 184x184x1x1]
%onnx::Conv_811[FLOAT, 552x184x1x1]
%onnx::Conv_812[FLOAT, 552]
%onnx::Conv_814[FLOAT, 552x1x3x3]
%onnx::Conv_817[FLOAT, 184x552x1x1]
%onnx::Conv_820[FLOAT, 184x92x1x1]
%onnx::Conv_823[FLOAT, 184x1x5x5]
%onnx::Conv_826[FLOAT, 184x92x1x1]
%onnx::Conv_829[FLOAT, 552x184x1x1]
%onnx::Conv_832[FLOAT, 552x1x3x3]
%onnx::Conv_835[FLOAT, 352x552x1x1]
%onnx::Conv_836[FLOAT, 352]
%onnx::Conv_838[FLOAT, 1504x352x1x1]
%onnx::Conv_839[FLOAT, 1504]
) {
%onnx::Conv_833 = Identity(%onnx::Conv_812)
%onnx::Conv_830 = Identity(%onnx::Conv_812)
%onnx::Conv_827 = Identity(%onnx::Conv_800)
%onnx::Conv_824 = Identity(%onnx::Conv_800)
%onnx::Conv_821 = Identity(%onnx::Conv_800)
%onnx::Conv_818 = Identity(%onnx::Conv_800)
%onnx::Conv_815 = Identity(%onnx::Conv_812)
%onnx::Conv_809 = Identity(%onnx::Conv_800)
%onnx::Conv_806 = Identity(%onnx::Conv_800)
%onnx::Conv_803 = Identity(%onnx::Conv_800)
%onnx::Conv_797 = Identity(%onnx::Conv_782)
%onnx::Conv_794 = Identity(%onnx::Conv_782)
%onnx::Conv_791 = Identity(%onnx::Conv_782)
%onnx::Conv_788 = Identity(%onnx::Conv_785)
%onnx::Conv_779 = Identity(%onnx::Conv_746)
%onnx::Conv_776 = Identity(%onnx::Conv_746)
%onnx::Conv_773 = Identity(%onnx::Conv_746)
%onnx::Conv_770 = Identity(%onnx::Conv_758)
%onnx::Conv_767 = Identity(%onnx::Conv_758)
%onnx::Conv_764 = Identity(%onnx::Conv_746)
%onnx::Conv_761 = Identity(%onnx::Conv_758)
%onnx::Conv_755 = Identity(%onnx::Conv_746)
%onnx::Conv_752 = Identity(%onnx::Conv_713)
%onnx::Conv_749 = Identity(%onnx::Conv_713)
%onnx::Conv_743 = Identity(%onnx::Conv_713)
%onnx::Conv_740 = Identity(%onnx::Conv_713)
%onnx::Conv_737 = Identity(%onnx::Conv_710)
%onnx::Conv_734 = Identity(%onnx::Conv_710)
%onnx::Conv_731 = Identity(%onnx::Conv_710)
%onnx::Conv_728 = Identity(%onnx::Conv_710)
%onnx::Conv_725 = Identity(%onnx::Conv_722)
%onnx::Conv_719 = Identity(%onnx::Conv_710)
%onnx::Conv_716 = Identity(%onnx::Conv_713)
%onnx::Conv_707 = Identity(%onnx::Conv_674)
%onnx::Conv_704 = Identity(%onnx::Conv_674)
%onnx::Conv_701 = Identity(%onnx::Conv_674)
%onnx::Conv_698 = Identity(%onnx::Conv_677)
%onnx::Conv_695 = Identity(%onnx::Conv_677)
%onnx::Conv_692 = Identity(%onnx::Conv_674)
%onnx::Conv_689 = Identity(%onnx::Conv_674)
%onnx::Conv_686 = Identity(%onnx::Conv_674)
%onnx::Conv_683 = Identity(%onnx::Conv_674)
%onnx::Conv_680 = Identity(%onnx::Conv_677)
%onnx::Conv_671 = Identity(%onnx::Conv_662)
%onnx::Conv_668 = Identity(%onnx::Conv_662)
%onnx::Conv_665 = Identity(%onnx::Conv_662)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_661, %onnx::Conv_662)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.1/relu/Relu_output_0 = Relu(%/cells.1/conv/Conv_output_0)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/relu/Relu_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/relu/Relu_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_823, %onnx::Conv_824)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_826, %onnx::Conv_827)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_829, %onnx::Conv_830)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_832, %onnx::Conv_833)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_835, %onnx::Conv_836)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_838, %onnx::Conv_839)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%659 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %659
}
|
val_accuracy
| 0
| 67,505,024
| 1,632,020
|
{'zcp_synflow': 70.89983396559502, 'zcp_zen': 63.74177932739258, 'zcp_epe_nas': 28.814151239893455, 'zcp_fisher': 0.10169743001461029, 'zcp_flops': 67505024.0, 'zcp_grad_norm': 24.513511657714844, 'zcp_grasp': -0.47586822509765625, 'zcp_jacov': -16.08144327529103, 'zcp_l2_norm': 586.2539672851562, 'zcp_nwot': 214.81478621654077, 'zcp_params': 1632020.0, 'zcp_plain': -0.007620286196470261, 'zcp_snip': 37.95701217651367, 'lat_1080ti_1': 0.5858581711498384, 'lat_1080ti_32': 0.6130271288360369, 'lat_1080ti_64': 0.5467568175621889, 'lat_2080ti_1': 0.5784719647336098, 'lat_2080ti_32': 0.6227549701811107, 'lat_2080ti_64': 0.5408799772730091, 'lat_essential_ph_1': 0.24528301886792453, 'lat_eyeriss': 0.4624906020207947, 'lat_fpga': 0.44921283606606044, 'lat_gold_6226': 0.31127905986261384, 'lat_gold_6240': 0.43246258785256264, 'lat_pixel2': 0.43478260869565216, 'lat_pixel3': 0.46258268751632614, 'lat_raspi4': 0.5071236654893815, 'lat_samsung_a50': 0.18947368421052632, 'lat_samsung_s7': 0.16535433070866143, 'lat_silver_4114': 0.47928943522345446, 'lat_silver_4210r': 0.5272080757408517, 'lat_titan_rtx_1': 0.5560153003742695, 'lat_titan_rtx_32': 0.595702589951272, 'lat_titan_rtx_64': 0.5711264799791534, 'lat_titanx_1': 0.2844464202925405, 'lat_titanx_32': 0.5686081649739225, 'lat_titanx_64': 0.5252240494601974, 'lat_titanxp_1': 0.515438533511534, 'lat_titanxp_32': 0.6076191436534177, 'lat_titanxp_64': 0.540146804187833}
| |
FBNet_2401
|
FBNet
|
2401
|
2401
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_622[FLOAT, 16x3x3x3]
%onnx::Conv_623[FLOAT, 16]
%onnx::Conv_625[FLOAT, 16x16x1x1]
%onnx::Conv_628[FLOAT, 16x1x3x3]
%onnx::Conv_631[FLOAT, 16x16x1x1]
%onnx::Conv_634[FLOAT, 96x16x1x1]
%onnx::Conv_635[FLOAT, 96]
%onnx::Conv_637[FLOAT, 96x1x3x3]
%onnx::Conv_640[FLOAT, 24x96x1x1]
%onnx::Conv_641[FLOAT, 24]
%onnx::Conv_643[FLOAT, 24x24x1x1]
%onnx::Conv_646[FLOAT, 24x1x5x5]
%onnx::Conv_649[FLOAT, 24x24x1x1]
%onnx::Conv_652[FLOAT, 24x24x1x1]
%onnx::Conv_655[FLOAT, 24x1x3x3]
%onnx::Conv_658[FLOAT, 24x24x1x1]
%onnx::Conv_661[FLOAT, 144x24x1x1]
%onnx::Conv_662[FLOAT, 144]
%onnx::Conv_664[FLOAT, 144x1x3x3]
%onnx::Conv_667[FLOAT, 24x144x1x1]
%onnx::Conv_670[FLOAT, 24x12x1x1]
%onnx::Conv_673[FLOAT, 24x1x5x5]
%onnx::Conv_676[FLOAT, 32x12x1x1]
%onnx::Conv_677[FLOAT, 32]
%onnx::Conv_679[FLOAT, 192x32x1x1]
%onnx::Conv_680[FLOAT, 192]
%onnx::Conv_682[FLOAT, 192x1x3x3]
%onnx::Conv_685[FLOAT, 32x192x1x1]
%onnx::Conv_688[FLOAT, 192x32x1x1]
%onnx::Conv_691[FLOAT, 192x1x5x5]
%onnx::Conv_694[FLOAT, 32x192x1x1]
%onnx::Conv_697[FLOAT, 32x16x1x1]
%onnx::Conv_700[FLOAT, 32x1x3x3]
%onnx::Conv_703[FLOAT, 32x16x1x1]
%onnx::Conv_706[FLOAT, 96x32x1x1]
%onnx::Conv_709[FLOAT, 96x1x3x3]
%onnx::Conv_712[FLOAT, 64x96x1x1]
%onnx::Conv_713[FLOAT, 64]
%onnx::Conv_715[FLOAT, 192x64x1x1]
%onnx::Conv_718[FLOAT, 192x1x5x5]
%onnx::Conv_721[FLOAT, 64x192x1x1]
%onnx::Conv_724[FLOAT, 64x32x1x1]
%onnx::Conv_727[FLOAT, 64x1x3x3]
%onnx::Conv_730[FLOAT, 64x32x1x1]
%onnx::Conv_733[FLOAT, 192x64x1x1]
%onnx::Conv_736[FLOAT, 192x1x5x5]
%onnx::Conv_739[FLOAT, 64x192x1x1]
%onnx::Conv_742[FLOAT, 192x64x1x1]
%onnx::Conv_745[FLOAT, 192x1x5x5]
%onnx::Conv_748[FLOAT, 112x192x1x1]
%onnx::Conv_749[FLOAT, 112]
%onnx::Conv_751[FLOAT, 336x112x1x1]
%onnx::Conv_752[FLOAT, 336]
%onnx::Conv_754[FLOAT, 336x1x3x3]
%onnx::Conv_757[FLOAT, 112x336x1x1]
%onnx::Conv_760[FLOAT, 336x112x1x1]
%onnx::Conv_763[FLOAT, 336x1x5x5]
%onnx::Conv_766[FLOAT, 112x336x1x1]
%onnx::Conv_769[FLOAT, 112x112x1x1]
%onnx::Conv_772[FLOAT, 112x1x3x3]
%onnx::Conv_775[FLOAT, 184x112x1x1]
%onnx::Conv_776[FLOAT, 184]
%onnx::Conv_778[FLOAT, 552x184x1x1]
%onnx::Conv_779[FLOAT, 552]
%onnx::Conv_781[FLOAT, 552x1x3x3]
%onnx::Conv_784[FLOAT, 184x552x1x1]
%onnx::Conv_787[FLOAT, 1104x184x1x1]
%onnx::Conv_788[FLOAT, 1104]
%onnx::Conv_790[FLOAT, 1104x1x3x3]
%onnx::Conv_793[FLOAT, 184x1104x1x1]
%onnx::Conv_796[FLOAT, 1104x184x1x1]
%onnx::Conv_799[FLOAT, 1104x1x5x5]
%onnx::Conv_802[FLOAT, 352x1104x1x1]
%onnx::Conv_803[FLOAT, 352]
%onnx::Conv_805[FLOAT, 1504x352x1x1]
%onnx::Conv_806[FLOAT, 1504]
) {
%onnx::Conv_800 = Identity(%onnx::Conv_788)
%onnx::Conv_797 = Identity(%onnx::Conv_788)
%onnx::Conv_794 = Identity(%onnx::Conv_776)
%onnx::Conv_791 = Identity(%onnx::Conv_788)
%onnx::Conv_785 = Identity(%onnx::Conv_776)
%onnx::Conv_782 = Identity(%onnx::Conv_779)
%onnx::Conv_773 = Identity(%onnx::Conv_749)
%onnx::Conv_770 = Identity(%onnx::Conv_749)
%onnx::Conv_767 = Identity(%onnx::Conv_749)
%onnx::Conv_764 = Identity(%onnx::Conv_752)
%onnx::Conv_761 = Identity(%onnx::Conv_752)
%onnx::Conv_758 = Identity(%onnx::Conv_749)
%onnx::Conv_755 = Identity(%onnx::Conv_752)
%onnx::Conv_746 = Identity(%onnx::Conv_680)
%onnx::Conv_743 = Identity(%onnx::Conv_680)
%onnx::Conv_740 = Identity(%onnx::Conv_713)
%onnx::Conv_737 = Identity(%onnx::Conv_680)
%onnx::Conv_734 = Identity(%onnx::Conv_680)
%onnx::Conv_731 = Identity(%onnx::Conv_713)
%onnx::Conv_728 = Identity(%onnx::Conv_713)
%onnx::Conv_725 = Identity(%onnx::Conv_713)
%onnx::Conv_722 = Identity(%onnx::Conv_713)
%onnx::Conv_719 = Identity(%onnx::Conv_680)
%onnx::Conv_716 = Identity(%onnx::Conv_680)
%onnx::Conv_710 = Identity(%onnx::Conv_635)
%onnx::Conv_707 = Identity(%onnx::Conv_635)
%onnx::Conv_704 = Identity(%onnx::Conv_677)
%onnx::Conv_701 = Identity(%onnx::Conv_677)
%onnx::Conv_698 = Identity(%onnx::Conv_677)
%onnx::Conv_695 = Identity(%onnx::Conv_677)
%onnx::Conv_692 = Identity(%onnx::Conv_680)
%onnx::Conv_689 = Identity(%onnx::Conv_680)
%onnx::Conv_686 = Identity(%onnx::Conv_677)
%onnx::Conv_683 = Identity(%onnx::Conv_680)
%onnx::Conv_674 = Identity(%onnx::Conv_641)
%onnx::Conv_671 = Identity(%onnx::Conv_641)
%onnx::Conv_668 = Identity(%onnx::Conv_641)
%onnx::Conv_665 = Identity(%onnx::Conv_662)
%onnx::Conv_659 = Identity(%onnx::Conv_641)
%onnx::Conv_656 = Identity(%onnx::Conv_641)
%onnx::Conv_653 = Identity(%onnx::Conv_641)
%onnx::Conv_650 = Identity(%onnx::Conv_641)
%onnx::Conv_647 = Identity(%onnx::Conv_641)
%onnx::Conv_644 = Identity(%onnx::Conv_641)
%onnx::Conv_638 = Identity(%onnx::Conv_635)
%onnx::Conv_632 = Identity(%onnx::Conv_623)
%onnx::Conv_629 = Identity(%onnx::Conv_623)
%onnx::Conv_626 = Identity(%onnx::Conv_623)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_622, %onnx::Conv_623)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_625, %onnx::Conv_626)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_628, %onnx::Conv_629)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_631, %onnx::Conv_632)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_634, %onnx::Conv_635)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_637, %onnx::Conv_638)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_640, %onnx::Conv_641)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_643, %onnx::Conv_644)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_646, %onnx::Conv_647)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_649, %onnx::Conv_650)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_652, %onnx::Conv_653)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_655, %onnx::Conv_656)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_658, %onnx::Conv_659)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%620 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %620
}
|
val_accuracy
| 0
| 77,602,688
| 2,309,188
|
{'zcp_synflow': 76.93476600816014, 'zcp_zen': 67.83760833740234, 'zcp_epe_nas': 10.333957291789718, 'zcp_fisher': 0.09264073520898819, 'zcp_flops': 77602688.0, 'zcp_grad_norm': 23.409168243408203, 'zcp_grasp': -0.19086837768554688, 'zcp_jacov': -16.06329787077916, 'zcp_l2_norm': 643.477783203125, 'zcp_nwot': 214.37383457546363, 'zcp_params': 2309188.0, 'zcp_plain': 0.006686722859740257, 'zcp_snip': 44.300872802734375, 'lat_1080ti_1': 0.623785192502496, 'lat_1080ti_32': 0.5554075131714757, 'lat_1080ti_64': 0.45137612046346864, 'lat_2080ti_1': 0.5880780724560999, 'lat_2080ti_32': 0.5351336428491261, 'lat_2080ti_64': 0.49162028442525796, 'lat_essential_ph_1': 0.22641509433962265, 'lat_eyeriss': 0.5496121614435296, 'lat_fpga': 0.6334418508000277, 'lat_gold_6226': 0.46990550197990977, 'lat_gold_6240': 0.6269160997451018, 'lat_pixel2': 0.5217391304347826, 'lat_pixel3': 0.5143477531108074, 'lat_raspi4': 0.598579146861213, 'lat_samsung_a50': 0.25263157894736843, 'lat_samsung_s7': 0.5984251968503937, 'lat_silver_4114': 0.6461550460208729, 'lat_silver_4210r': 0.6810173541867199, 'lat_titan_rtx_1': 0.5460669398958705, 'lat_titan_rtx_32': 0.517009668518206, 'lat_titan_rtx_64': 0.4975589676066239, 'lat_titanx_1': 0.29516615969827703, 'lat_titanx_32': 0.4888800600417003, 'lat_titanx_64': 0.4484508070855101, 'lat_titanxp_1': 0.5216389541509933, 'lat_titanxp_32': 0.5205775047111613, 'lat_titanxp_64': 0.46604318981995}
| |
FBNet_1025
|
FBNet
|
1025
|
1025
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_715[FLOAT, 16x3x3x3]
%onnx::Conv_716[FLOAT, 16]
%onnx::Conv_718[FLOAT, 16x8x1x1]
%onnx::Conv_721[FLOAT, 16x1x5x5]
%onnx::Conv_724[FLOAT, 16x8x1x1]
%onnx::Conv_727[FLOAT, 16x8x1x1]
%onnx::Conv_730[FLOAT, 16x1x5x5]
%onnx::Conv_733[FLOAT, 24x8x1x1]
%onnx::Conv_734[FLOAT, 24]
%onnx::Conv_736[FLOAT, 24x24x1x1]
%onnx::Conv_739[FLOAT, 24x1x3x3]
%onnx::Conv_742[FLOAT, 24x24x1x1]
%onnx::Conv_745[FLOAT, 24x12x1x1]
%onnx::Conv_748[FLOAT, 24x1x5x5]
%onnx::Conv_751[FLOAT, 24x12x1x1]
%onnx::Conv_754[FLOAT, 72x24x1x1]
%onnx::Conv_755[FLOAT, 72]
%onnx::Conv_757[FLOAT, 72x1x3x3]
%onnx::Conv_760[FLOAT, 32x72x1x1]
%onnx::Conv_761[FLOAT, 32]
%onnx::Conv_763[FLOAT, 192x32x1x1]
%onnx::Conv_764[FLOAT, 192]
%onnx::Conv_766[FLOAT, 192x1x3x3]
%onnx::Conv_769[FLOAT, 32x192x1x1]
%onnx::Conv_772[FLOAT, 32x16x1x1]
%onnx::Conv_775[FLOAT, 32x1x3x3]
%onnx::Conv_778[FLOAT, 32x16x1x1]
%onnx::Conv_781[FLOAT, 32x32x1x1]
%onnx::Conv_784[FLOAT, 32x1x5x5]
%onnx::Conv_787[FLOAT, 64x32x1x1]
%onnx::Conv_788[FLOAT, 64]
%onnx::Conv_790[FLOAT, 192x64x1x1]
%onnx::Conv_793[FLOAT, 192x1x3x3]
%onnx::Conv_796[FLOAT, 64x192x1x1]
%onnx::Conv_799[FLOAT, 64x32x1x1]
%onnx::Conv_802[FLOAT, 64x1x5x5]
%onnx::Conv_805[FLOAT, 64x32x1x1]
%onnx::Conv_808[FLOAT, 192x64x1x1]
%onnx::Conv_811[FLOAT, 192x1x3x3]
%onnx::Conv_814[FLOAT, 64x192x1x1]
%onnx::Conv_817[FLOAT, 64x64x1x1]
%onnx::Conv_820[FLOAT, 64x1x3x3]
%onnx::Conv_823[FLOAT, 112x64x1x1]
%onnx::Conv_824[FLOAT, 112]
%onnx::Conv_826[FLOAT, 112x56x1x1]
%onnx::Conv_829[FLOAT, 112x1x3x3]
%onnx::Conv_832[FLOAT, 112x56x1x1]
%onnx::Conv_835[FLOAT, 112x112x1x1]
%onnx::Conv_838[FLOAT, 112x1x5x5]
%onnx::Conv_841[FLOAT, 112x112x1x1]
%onnx::Conv_844[FLOAT, 672x112x1x1]
%onnx::Conv_845[FLOAT, 672]
%onnx::Conv_847[FLOAT, 672x1x5x5]
%onnx::Conv_850[FLOAT, 112x672x1x1]
%onnx::Conv_853[FLOAT, 112x112x1x1]
%onnx::Conv_856[FLOAT, 112x1x3x3]
%onnx::Conv_859[FLOAT, 184x112x1x1]
%onnx::Conv_860[FLOAT, 184]
%onnx::Conv_862[FLOAT, 184x184x1x1]
%onnx::Conv_865[FLOAT, 184x1x5x5]
%onnx::Conv_868[FLOAT, 184x184x1x1]
%onnx::Conv_871[FLOAT, 184x92x1x1]
%onnx::Conv_874[FLOAT, 184x1x5x5]
%onnx::Conv_877[FLOAT, 184x92x1x1]
%onnx::Conv_880[FLOAT, 1104x184x1x1]
%onnx::Conv_881[FLOAT, 1104]
%onnx::Conv_883[FLOAT, 1104x1x3x3]
%onnx::Conv_886[FLOAT, 184x1104x1x1]
%onnx::Conv_889[FLOAT, 184x92x1x1]
%onnx::Conv_892[FLOAT, 184x1x3x3]
%onnx::Conv_895[FLOAT, 352x92x1x1]
%onnx::Conv_896[FLOAT, 352]
%onnx::Conv_898[FLOAT, 1504x352x1x1]
%onnx::Conv_899[FLOAT, 1504]
) {
%onnx::Conv_893 = Identity(%onnx::Conv_860)
%onnx::Conv_890 = Identity(%onnx::Conv_860)
%onnx::Conv_887 = Identity(%onnx::Conv_860)
%onnx::Conv_884 = Identity(%onnx::Conv_881)
%onnx::Conv_878 = Identity(%onnx::Conv_860)
%onnx::Conv_875 = Identity(%onnx::Conv_860)
%onnx::Conv_872 = Identity(%onnx::Conv_860)
%onnx::Conv_869 = Identity(%onnx::Conv_860)
%onnx::Conv_866 = Identity(%onnx::Conv_860)
%onnx::Conv_863 = Identity(%onnx::Conv_860)
%onnx::Conv_857 = Identity(%onnx::Conv_824)
%onnx::Conv_854 = Identity(%onnx::Conv_824)
%onnx::Conv_851 = Identity(%onnx::Conv_824)
%onnx::Conv_848 = Identity(%onnx::Conv_845)
%onnx::Conv_842 = Identity(%onnx::Conv_824)
%onnx::Conv_839 = Identity(%onnx::Conv_824)
%onnx::Conv_836 = Identity(%onnx::Conv_824)
%onnx::Conv_833 = Identity(%onnx::Conv_824)
%onnx::Conv_830 = Identity(%onnx::Conv_824)
%onnx::Conv_827 = Identity(%onnx::Conv_824)
%onnx::Conv_821 = Identity(%onnx::Conv_788)
%onnx::Conv_818 = Identity(%onnx::Conv_788)
%onnx::Conv_815 = Identity(%onnx::Conv_788)
%onnx::Conv_812 = Identity(%onnx::Conv_764)
%onnx::Conv_809 = Identity(%onnx::Conv_764)
%onnx::Conv_806 = Identity(%onnx::Conv_788)
%onnx::Conv_803 = Identity(%onnx::Conv_788)
%onnx::Conv_800 = Identity(%onnx::Conv_788)
%onnx::Conv_797 = Identity(%onnx::Conv_788)
%onnx::Conv_794 = Identity(%onnx::Conv_764)
%onnx::Conv_791 = Identity(%onnx::Conv_764)
%onnx::Conv_785 = Identity(%onnx::Conv_761)
%onnx::Conv_782 = Identity(%onnx::Conv_761)
%onnx::Conv_779 = Identity(%onnx::Conv_761)
%onnx::Conv_776 = Identity(%onnx::Conv_761)
%onnx::Conv_773 = Identity(%onnx::Conv_761)
%onnx::Conv_770 = Identity(%onnx::Conv_761)
%onnx::Conv_767 = Identity(%onnx::Conv_764)
%onnx::Conv_758 = Identity(%onnx::Conv_755)
%onnx::Conv_752 = Identity(%onnx::Conv_734)
%onnx::Conv_749 = Identity(%onnx::Conv_734)
%onnx::Conv_746 = Identity(%onnx::Conv_734)
%onnx::Conv_743 = Identity(%onnx::Conv_734)
%onnx::Conv_740 = Identity(%onnx::Conv_734)
%onnx::Conv_737 = Identity(%onnx::Conv_734)
%onnx::Conv_731 = Identity(%onnx::Conv_716)
%onnx::Conv_728 = Identity(%onnx::Conv_716)
%onnx::Conv_725 = Identity(%onnx::Conv_716)
%onnx::Conv_722 = Identity(%onnx::Conv_716)
%onnx::Conv_719 = Identity(%onnx::Conv_716)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_715, %onnx::Conv_716)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_823, %onnx::Conv_824)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_826, %onnx::Conv_827)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_829, %onnx::Conv_830)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_832, %onnx::Conv_833)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_835, %onnx::Conv_836)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_838, %onnx::Conv_839)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_841, %onnx::Conv_842)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_844, %onnx::Conv_845)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_847, %onnx::Conv_848)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_850, %onnx::Conv_851)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_853, %onnx::Conv_854)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_856, %onnx::Conv_857)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_859, %onnx::Conv_860)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_862, %onnx::Conv_863)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_865, %onnx::Conv_866)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_868, %onnx::Conv_869)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_871, %onnx::Conv_872)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_874, %onnx::Conv_875)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_877, %onnx::Conv_878)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_880, %onnx::Conv_881)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_883, %onnx::Conv_884)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_886, %onnx::Conv_887)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_889, %onnx::Conv_890)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_892, %onnx::Conv_893)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_895, %onnx::Conv_896)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_898, %onnx::Conv_899)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%713 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %713
}
|
val_accuracy
| 0
| 50,241,152
| 1,620,484
|
{'zcp_synflow': 71.92326830085572, 'zcp_zen': 63.706207275390625, 'zcp_epe_nas': 12.014340357654154, 'zcp_fisher': 0.10795985907316208, 'zcp_flops': 50241152.0, 'zcp_grad_norm': 23.66715431213379, 'zcp_grasp': 0.004845619201660156, 'zcp_jacov': -16.072310667408217, 'zcp_l2_norm': 561.901123046875, 'zcp_nwot': 205.04850455604196, 'zcp_params': 1620484.0, 'zcp_plain': 0.0019517767941579223, 'zcp_snip': 35.24565124511719, 'lat_1080ti_1': 0.6248989332183194, 'lat_1080ti_32': 0.4784781193622443, 'lat_1080ti_64': 0.26188050951841796, 'lat_2080ti_1': 0.6768182308063494, 'lat_2080ti_32': 0.5116062308133867, 'lat_2080ti_64': 0.30701428561346517, 'lat_essential_ph_1': 0.18867924528301888, 'lat_eyeriss': 0.21303614325270928, 'lat_fpga': 0.2568400637251506, 'lat_gold_6226': 0.21057986098763048, 'lat_gold_6240': 0.4109251358514939, 'lat_pixel2': 0.4782608695652174, 'lat_pixel3': 0.2286235368052207, 'lat_raspi4': 0.26715760785103343, 'lat_samsung_a50': 0.2, 'lat_samsung_s7': 0.11811023622047244, 'lat_silver_4114': 0.6166313971470672, 'lat_silver_4210r': 0.4581221669479905, 'lat_titan_rtx_1': 0.6325473653273418, 'lat_titan_rtx_32': 0.5352059402873602, 'lat_titan_rtx_64': 0.3510946441481486, 'lat_titanx_1': 0.3419593184408099, 'lat_titanx_32': 0.4008468689890453, 'lat_titanx_64': 0.2776221852692277, 'lat_titanxp_1': 0.5937682599255956, 'lat_titanxp_32': 0.4589941414809199, 'lat_titanxp_64': 0.30780420207060083}
| |
FBNet_103
|
FBNet
|
103
|
103
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_643[FLOAT, 16x3x3x3]
%onnx::Conv_644[FLOAT, 16]
%onnx::Conv_646[FLOAT, 16x8x1x1]
%onnx::Conv_649[FLOAT, 16x1x3x3]
%onnx::Conv_652[FLOAT, 16x8x1x1]
%onnx::Conv_655[FLOAT, 16x8x1x1]
%onnx::Conv_658[FLOAT, 16x1x3x3]
%onnx::Conv_661[FLOAT, 24x8x1x1]
%onnx::Conv_662[FLOAT, 24]
%onnx::Conv_664[FLOAT, 24x24x1x1]
%onnx::Conv_667[FLOAT, 24x1x3x3]
%onnx::Conv_670[FLOAT, 24x24x1x1]
%onnx::Conv_673[FLOAT, 24x12x1x1]
%onnx::Conv_676[FLOAT, 24x1x3x3]
%onnx::Conv_679[FLOAT, 24x12x1x1]
%onnx::Conv_682[FLOAT, 24x12x1x1]
%onnx::Conv_685[FLOAT, 24x1x3x3]
%onnx::Conv_688[FLOAT, 32x12x1x1]
%onnx::Conv_689[FLOAT, 32]
%onnx::Conv_691[FLOAT, 32x32x1x1]
%onnx::Conv_694[FLOAT, 32x1x5x5]
%onnx::Conv_697[FLOAT, 32x32x1x1]
%onnx::Conv_700[FLOAT, 32x32x1x1]
%onnx::Conv_703[FLOAT, 32x1x3x3]
%onnx::Conv_706[FLOAT, 32x32x1x1]
%onnx::Conv_709[FLOAT, 32x16x1x1]
%onnx::Conv_712[FLOAT, 32x1x3x3]
%onnx::Conv_715[FLOAT, 32x16x1x1]
%onnx::Conv_718[FLOAT, 32x32x1x1]
%onnx::Conv_721[FLOAT, 32x1x3x3]
%onnx::Conv_724[FLOAT, 64x32x1x1]
%onnx::Conv_725[FLOAT, 64]
%onnx::Conv_727[FLOAT, 384x64x1x1]
%onnx::Conv_728[FLOAT, 384]
%onnx::Conv_730[FLOAT, 384x1x5x5]
%onnx::Conv_733[FLOAT, 64x384x1x1]
%onnx::Conv_736[FLOAT, 64x32x1x1]
%onnx::Conv_739[FLOAT, 64x1x5x5]
%onnx::Conv_742[FLOAT, 64x32x1x1]
%onnx::Conv_745[FLOAT, 64x64x1x1]
%onnx::Conv_748[FLOAT, 64x1x3x3]
%onnx::Conv_751[FLOAT, 64x64x1x1]
%onnx::Conv_754[FLOAT, 64x64x1x1]
%onnx::Conv_757[FLOAT, 64x1x5x5]
%onnx::Conv_760[FLOAT, 112x64x1x1]
%onnx::Conv_761[FLOAT, 112]
%onnx::Conv_763[FLOAT, 672x112x1x1]
%onnx::Conv_764[FLOAT, 672]
%onnx::Conv_766[FLOAT, 672x1x3x3]
%onnx::Conv_769[FLOAT, 112x672x1x1]
%onnx::Conv_772[FLOAT, 672x112x1x1]
%onnx::Conv_775[FLOAT, 672x1x5x5]
%onnx::Conv_778[FLOAT, 112x672x1x1]
%onnx::Conv_781[FLOAT, 184x112x1x1]
%onnx::Conv_782[FLOAT, 184]
%onnx::Conv_784[FLOAT, 184x184x1x1]
%onnx::Conv_787[FLOAT, 184x1x5x5]
%onnx::Conv_790[FLOAT, 184x184x1x1]
%onnx::Conv_793[FLOAT, 552x184x1x1]
%onnx::Conv_794[FLOAT, 552]
%onnx::Conv_796[FLOAT, 552x1x3x3]
%onnx::Conv_799[FLOAT, 184x552x1x1]
%onnx::Conv_802[FLOAT, 184x184x1x1]
%onnx::Conv_805[FLOAT, 184x1x5x5]
%onnx::Conv_808[FLOAT, 184x184x1x1]
%onnx::Conv_811[FLOAT, 352x184x1x1]
%onnx::Conv_812[FLOAT, 352]
%onnx::Conv_814[FLOAT, 1504x352x1x1]
%onnx::Conv_815[FLOAT, 1504]
) {
%onnx::Conv_809 = Identity(%onnx::Conv_782)
%onnx::Conv_806 = Identity(%onnx::Conv_782)
%onnx::Conv_803 = Identity(%onnx::Conv_782)
%onnx::Conv_800 = Identity(%onnx::Conv_782)
%onnx::Conv_797 = Identity(%onnx::Conv_794)
%onnx::Conv_791 = Identity(%onnx::Conv_782)
%onnx::Conv_788 = Identity(%onnx::Conv_782)
%onnx::Conv_785 = Identity(%onnx::Conv_782)
%onnx::Conv_779 = Identity(%onnx::Conv_761)
%onnx::Conv_776 = Identity(%onnx::Conv_764)
%onnx::Conv_773 = Identity(%onnx::Conv_764)
%onnx::Conv_770 = Identity(%onnx::Conv_761)
%onnx::Conv_767 = Identity(%onnx::Conv_764)
%onnx::Conv_758 = Identity(%onnx::Conv_725)
%onnx::Conv_755 = Identity(%onnx::Conv_725)
%onnx::Conv_752 = Identity(%onnx::Conv_725)
%onnx::Conv_749 = Identity(%onnx::Conv_725)
%onnx::Conv_746 = Identity(%onnx::Conv_725)
%onnx::Conv_743 = Identity(%onnx::Conv_725)
%onnx::Conv_740 = Identity(%onnx::Conv_725)
%onnx::Conv_737 = Identity(%onnx::Conv_725)
%onnx::Conv_734 = Identity(%onnx::Conv_725)
%onnx::Conv_731 = Identity(%onnx::Conv_728)
%onnx::Conv_722 = Identity(%onnx::Conv_689)
%onnx::Conv_719 = Identity(%onnx::Conv_689)
%onnx::Conv_716 = Identity(%onnx::Conv_689)
%onnx::Conv_713 = Identity(%onnx::Conv_689)
%onnx::Conv_710 = Identity(%onnx::Conv_689)
%onnx::Conv_707 = Identity(%onnx::Conv_689)
%onnx::Conv_704 = Identity(%onnx::Conv_689)
%onnx::Conv_701 = Identity(%onnx::Conv_689)
%onnx::Conv_698 = Identity(%onnx::Conv_689)
%onnx::Conv_695 = Identity(%onnx::Conv_689)
%onnx::Conv_692 = Identity(%onnx::Conv_689)
%onnx::Conv_686 = Identity(%onnx::Conv_662)
%onnx::Conv_683 = Identity(%onnx::Conv_662)
%onnx::Conv_680 = Identity(%onnx::Conv_662)
%onnx::Conv_677 = Identity(%onnx::Conv_662)
%onnx::Conv_674 = Identity(%onnx::Conv_662)
%onnx::Conv_671 = Identity(%onnx::Conv_662)
%onnx::Conv_668 = Identity(%onnx::Conv_662)
%onnx::Conv_665 = Identity(%onnx::Conv_662)
%onnx::Conv_659 = Identity(%onnx::Conv_644)
%onnx::Conv_656 = Identity(%onnx::Conv_644)
%onnx::Conv_653 = Identity(%onnx::Conv_644)
%onnx::Conv_650 = Identity(%onnx::Conv_644)
%onnx::Conv_647 = Identity(%onnx::Conv_644)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_643, %onnx::Conv_644)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_646, %onnx::Conv_647)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_649, %onnx::Conv_650)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_652, %onnx::Conv_653)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_655, %onnx::Conv_656)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_658, %onnx::Conv_659)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.21/relu/Relu_output_0 = Relu(%/cells.21/conv/Conv_output_0)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/relu/Relu_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%641 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %641
}
|
val_accuracy
| 0
| 49,752,576
| 1,561,316
|
{'zcp_synflow': 70.26932402812301, 'zcp_zen': 58.438194274902344, 'zcp_epe_nas': 19.98428080005745, 'zcp_fisher': 0.05177943781018257, 'zcp_flops': 49752576.0, 'zcp_grad_norm': 17.003141403198242, 'zcp_grasp': -0.027527809143066406, 'zcp_jacov': -16.05339031586395, 'zcp_l2_norm': 530.736083984375, 'zcp_nwot': 202.62758915132733, 'zcp_params': 1561316.0, 'zcp_plain': 0.005872313864529133, 'zcp_snip': 26.111366271972656, 'lat_1080ti_1': 0.4775750588539775, 'lat_1080ti_32': 0.35425884742498065, 'lat_1080ti_64': 0.16953224203661282, 'lat_2080ti_1': 0.49701012932095684, 'lat_2080ti_32': 0.3549305026726554, 'lat_2080ti_64': 0.1994630865866888, 'lat_essential_ph_1': 0.2830188679245283, 'lat_eyeriss': 0.16916362569453364, 'lat_fpga': 0.24966603667164722, 'lat_gold_6226': 0.20166546010238923, 'lat_gold_6240': 0.3285225670679963, 'lat_pixel2': 0.13043478260869565, 'lat_pixel3': 0.202590582171922, 'lat_raspi4': 0.20358054921287042, 'lat_samsung_a50': 0.08421052631578947, 'lat_samsung_s7': 0.08661417322834646, 'lat_silver_4114': 0.38384014044240544, 'lat_silver_4210r': 0.3539846110073291, 'lat_titan_rtx_1': 0.4768766061590806, 'lat_titan_rtx_32': 0.38054861100357334, 'lat_titan_rtx_64': 0.23858682627481168, 'lat_titanx_1': 0.25361044507994457, 'lat_titanx_32': 0.2769467736424277, 'lat_titanx_64': 0.16870990637109082, 'lat_titanxp_1': 0.4576976044806042, 'lat_titanxp_32': 0.3348504588774871, 'lat_titanxp_64': 0.20227116232407943}
| |
FBNet_1553
|
FBNet
|
1553
|
1553
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_714[FLOAT, 16x3x3x3]
%onnx::Conv_715[FLOAT, 16]
%onnx::Conv_717[FLOAT, 96x16x1x1]
%onnx::Conv_718[FLOAT, 96]
%onnx::Conv_720[FLOAT, 96x1x3x3]
%onnx::Conv_723[FLOAT, 16x96x1x1]
%onnx::Conv_726[FLOAT, 16x16x1x1]
%onnx::Conv_729[FLOAT, 16x1x3x3]
%onnx::Conv_732[FLOAT, 24x16x1x1]
%onnx::Conv_733[FLOAT, 24]
%onnx::Conv_735[FLOAT, 24x12x1x1]
%onnx::Conv_738[FLOAT, 24x1x3x3]
%onnx::Conv_741[FLOAT, 24x12x1x1]
%onnx::Conv_744[FLOAT, 24x12x1x1]
%onnx::Conv_747[FLOAT, 24x1x3x3]
%onnx::Conv_750[FLOAT, 24x12x1x1]
%onnx::Conv_753[FLOAT, 72x24x1x1]
%onnx::Conv_754[FLOAT, 72]
%onnx::Conv_756[FLOAT, 72x1x5x5]
%onnx::Conv_759[FLOAT, 24x72x1x1]
%onnx::Conv_762[FLOAT, 144x24x1x1]
%onnx::Conv_763[FLOAT, 144]
%onnx::Conv_765[FLOAT, 144x1x5x5]
%onnx::Conv_768[FLOAT, 32x144x1x1]
%onnx::Conv_769[FLOAT, 32]
%onnx::Conv_771[FLOAT, 192x32x1x1]
%onnx::Conv_772[FLOAT, 192]
%onnx::Conv_774[FLOAT, 192x1x5x5]
%onnx::Conv_777[FLOAT, 32x192x1x1]
%onnx::Conv_780[FLOAT, 96x32x1x1]
%onnx::Conv_783[FLOAT, 96x1x5x5]
%onnx::Conv_786[FLOAT, 32x96x1x1]
%onnx::Conv_789[FLOAT, 96x32x1x1]
%onnx::Conv_792[FLOAT, 96x1x5x5]
%onnx::Conv_795[FLOAT, 32x96x1x1]
%onnx::Conv_798[FLOAT, 192x32x1x1]
%onnx::Conv_801[FLOAT, 192x1x3x3]
%onnx::Conv_804[FLOAT, 64x192x1x1]
%onnx::Conv_805[FLOAT, 64]
%onnx::Conv_807[FLOAT, 64x32x1x1]
%onnx::Conv_810[FLOAT, 64x1x5x5]
%onnx::Conv_813[FLOAT, 64x32x1x1]
%onnx::Conv_816[FLOAT, 64x32x1x1]
%onnx::Conv_819[FLOAT, 64x1x5x5]
%onnx::Conv_822[FLOAT, 64x32x1x1]
%onnx::Conv_825[FLOAT, 192x64x1x1]
%onnx::Conv_828[FLOAT, 192x1x3x3]
%onnx::Conv_831[FLOAT, 64x192x1x1]
%onnx::Conv_834[FLOAT, 64x32x1x1]
%onnx::Conv_837[FLOAT, 64x1x3x3]
%onnx::Conv_840[FLOAT, 112x32x1x1]
%onnx::Conv_841[FLOAT, 112]
%onnx::Conv_843[FLOAT, 112x56x1x1]
%onnx::Conv_846[FLOAT, 112x1x5x5]
%onnx::Conv_849[FLOAT, 112x56x1x1]
%onnx::Conv_852[FLOAT, 112x56x1x1]
%onnx::Conv_855[FLOAT, 112x1x3x3]
%onnx::Conv_858[FLOAT, 112x56x1x1]
%onnx::Conv_861[FLOAT, 112x112x1x1]
%onnx::Conv_864[FLOAT, 112x1x5x5]
%onnx::Conv_867[FLOAT, 112x112x1x1]
%onnx::Conv_870[FLOAT, 184x112x1x1]
%onnx::Conv_871[FLOAT, 184]
%onnx::Conv_873[FLOAT, 552x184x1x1]
%onnx::Conv_874[FLOAT, 552]
%onnx::Conv_876[FLOAT, 552x1x3x3]
%onnx::Conv_879[FLOAT, 184x552x1x1]
%onnx::Conv_882[FLOAT, 552x184x1x1]
%onnx::Conv_885[FLOAT, 552x1x3x3]
%onnx::Conv_888[FLOAT, 184x552x1x1]
%onnx::Conv_891[FLOAT, 1104x184x1x1]
%onnx::Conv_892[FLOAT, 1104]
%onnx::Conv_894[FLOAT, 1104x1x3x3]
%onnx::Conv_897[FLOAT, 184x1104x1x1]
%onnx::Conv_900[FLOAT, 352x184x1x1]
%onnx::Conv_901[FLOAT, 352]
%onnx::Conv_903[FLOAT, 1504x352x1x1]
%onnx::Conv_904[FLOAT, 1504]
) {
%onnx::Conv_898 = Identity(%onnx::Conv_871)
%onnx::Conv_895 = Identity(%onnx::Conv_892)
%onnx::Conv_889 = Identity(%onnx::Conv_871)
%onnx::Conv_886 = Identity(%onnx::Conv_874)
%onnx::Conv_883 = Identity(%onnx::Conv_874)
%onnx::Conv_880 = Identity(%onnx::Conv_871)
%onnx::Conv_877 = Identity(%onnx::Conv_874)
%onnx::Conv_868 = Identity(%onnx::Conv_841)
%onnx::Conv_865 = Identity(%onnx::Conv_841)
%onnx::Conv_862 = Identity(%onnx::Conv_841)
%onnx::Conv_859 = Identity(%onnx::Conv_841)
%onnx::Conv_856 = Identity(%onnx::Conv_841)
%onnx::Conv_853 = Identity(%onnx::Conv_841)
%onnx::Conv_850 = Identity(%onnx::Conv_841)
%onnx::Conv_847 = Identity(%onnx::Conv_841)
%onnx::Conv_844 = Identity(%onnx::Conv_841)
%onnx::Conv_838 = Identity(%onnx::Conv_805)
%onnx::Conv_835 = Identity(%onnx::Conv_805)
%onnx::Conv_832 = Identity(%onnx::Conv_805)
%onnx::Conv_829 = Identity(%onnx::Conv_772)
%onnx::Conv_826 = Identity(%onnx::Conv_772)
%onnx::Conv_823 = Identity(%onnx::Conv_805)
%onnx::Conv_820 = Identity(%onnx::Conv_805)
%onnx::Conv_817 = Identity(%onnx::Conv_805)
%onnx::Conv_814 = Identity(%onnx::Conv_805)
%onnx::Conv_811 = Identity(%onnx::Conv_805)
%onnx::Conv_808 = Identity(%onnx::Conv_805)
%onnx::Conv_802 = Identity(%onnx::Conv_772)
%onnx::Conv_799 = Identity(%onnx::Conv_772)
%onnx::Conv_796 = Identity(%onnx::Conv_769)
%onnx::Conv_793 = Identity(%onnx::Conv_718)
%onnx::Conv_790 = Identity(%onnx::Conv_718)
%onnx::Conv_787 = Identity(%onnx::Conv_769)
%onnx::Conv_784 = Identity(%onnx::Conv_718)
%onnx::Conv_781 = Identity(%onnx::Conv_718)
%onnx::Conv_778 = Identity(%onnx::Conv_769)
%onnx::Conv_775 = Identity(%onnx::Conv_772)
%onnx::Conv_766 = Identity(%onnx::Conv_763)
%onnx::Conv_760 = Identity(%onnx::Conv_733)
%onnx::Conv_757 = Identity(%onnx::Conv_754)
%onnx::Conv_751 = Identity(%onnx::Conv_733)
%onnx::Conv_748 = Identity(%onnx::Conv_733)
%onnx::Conv_745 = Identity(%onnx::Conv_733)
%onnx::Conv_742 = Identity(%onnx::Conv_733)
%onnx::Conv_739 = Identity(%onnx::Conv_733)
%onnx::Conv_736 = Identity(%onnx::Conv_733)
%onnx::Conv_730 = Identity(%onnx::Conv_715)
%onnx::Conv_727 = Identity(%onnx::Conv_715)
%onnx::Conv_724 = Identity(%onnx::Conv_715)
%onnx::Conv_721 = Identity(%onnx::Conv_718)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_714, %onnx::Conv_715)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_858, %onnx::Conv_859)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_861, %onnx::Conv_862)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_864, %onnx::Conv_865)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_867, %onnx::Conv_868)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_870, %onnx::Conv_871)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_873, %onnx::Conv_874)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_876, %onnx::Conv_877)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_879, %onnx::Conv_880)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_882, %onnx::Conv_883)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_885, %onnx::Conv_886)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_888, %onnx::Conv_889)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_891, %onnx::Conv_892)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_894, %onnx::Conv_895)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_897, %onnx::Conv_898)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_900, %onnx::Conv_901)
%/cells.21/relu/Relu_output_0 = Relu(%/cells.21/conv/Conv_output_0)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/relu/Relu_output_0, %onnx::Conv_903, %onnx::Conv_904)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%712 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %712
}
|
val_accuracy
| 0
| 62,089,472
| 1,799,004
|
{'zcp_synflow': 77.67362248047358, 'zcp_zen': 67.21208953857422, 'zcp_epe_nas': 9.54543548722008, 'zcp_fisher': 0.14088934659957886, 'zcp_flops': 62089472.0, 'zcp_grad_norm': 25.210172653198242, 'zcp_grasp': -0.024648666381835938, 'zcp_jacov': -16.05307319021746, 'zcp_l2_norm': 605.482177734375, 'zcp_nwot': 213.57434807385027, 'zcp_params': 1799004.0, 'zcp_plain': 0.002713225781917572, 'zcp_snip': 41.45919418334961, 'lat_1080ti_1': 0.7488350078016554, 'lat_1080ti_32': 0.5943520486289847, 'lat_1080ti_64': 0.5217848786315787, 'lat_2080ti_1': 0.7493940760276434, 'lat_2080ti_32': 0.6343517265493509, 'lat_2080ti_64': 0.5320873729622049, 'lat_essential_ph_1': 0.41509433962264153, 'lat_eyeriss': 0.4306474978453413, 'lat_fpga': 0.3365855589309215, 'lat_gold_6226': 0.31572732730598685, 'lat_gold_6240': 0.5123204710769487, 'lat_pixel2': 0.2826086956521739, 'lat_pixel3': 0.4168034921854682, 'lat_raspi4': 0.43418345839155315, 'lat_samsung_a50': 0.17894736842105263, 'lat_samsung_s7': 0.2125984251968504, 'lat_silver_4114': 0.5275354197959828, 'lat_silver_4210r': 0.5273580034251846, 'lat_titan_rtx_1': 0.7071745693212794, 'lat_titan_rtx_32': 0.6267476454451655, 'lat_titan_rtx_64': 0.5793247065366764, 'lat_titanx_1': 0.3744387403501526, 'lat_titanx_32': 0.6115607927946762, 'lat_titanx_64': 0.5055589460018948, 'lat_titanxp_1': 0.6726733350158363, 'lat_titanxp_32': 0.6154221571744285, 'lat_titanxp_64': 0.5164330290906799}
| |
FBNet_4471
|
FBNet
|
4471
|
4471
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_687[FLOAT, 16x3x3x3]
%onnx::Conv_688[FLOAT, 16]
%onnx::Conv_690[FLOAT, 16x16x1x1]
%onnx::Conv_693[FLOAT, 16x1x3x3]
%onnx::Conv_696[FLOAT, 16x16x1x1]
%onnx::Conv_699[FLOAT, 96x16x1x1]
%onnx::Conv_700[FLOAT, 96]
%onnx::Conv_702[FLOAT, 96x1x5x5]
%onnx::Conv_705[FLOAT, 24x96x1x1]
%onnx::Conv_706[FLOAT, 24]
%onnx::Conv_708[FLOAT, 72x24x1x1]
%onnx::Conv_709[FLOAT, 72]
%onnx::Conv_711[FLOAT, 72x1x5x5]
%onnx::Conv_714[FLOAT, 24x72x1x1]
%onnx::Conv_717[FLOAT, 24x12x1x1]
%onnx::Conv_720[FLOAT, 24x1x3x3]
%onnx::Conv_723[FLOAT, 24x12x1x1]
%onnx::Conv_726[FLOAT, 24x24x1x1]
%onnx::Conv_729[FLOAT, 24x1x3x3]
%onnx::Conv_732[FLOAT, 24x24x1x1]
%onnx::Conv_735[FLOAT, 144x24x1x1]
%onnx::Conv_736[FLOAT, 144]
%onnx::Conv_738[FLOAT, 144x1x3x3]
%onnx::Conv_741[FLOAT, 32x144x1x1]
%onnx::Conv_742[FLOAT, 32]
%onnx::Conv_744[FLOAT, 192x32x1x1]
%onnx::Conv_745[FLOAT, 192]
%onnx::Conv_747[FLOAT, 192x1x5x5]
%onnx::Conv_750[FLOAT, 32x192x1x1]
%onnx::Conv_753[FLOAT, 32x32x1x1]
%onnx::Conv_756[FLOAT, 32x1x3x3]
%onnx::Conv_759[FLOAT, 32x32x1x1]
%onnx::Conv_762[FLOAT, 32x32x1x1]
%onnx::Conv_765[FLOAT, 32x1x3x3]
%onnx::Conv_768[FLOAT, 32x32x1x1]
%onnx::Conv_771[FLOAT, 64x32x1x1]
%onnx::Conv_772[FLOAT, 64]
%onnx::Conv_774[FLOAT, 64x32x1x1]
%onnx::Conv_777[FLOAT, 64x1x3x3]
%onnx::Conv_780[FLOAT, 64x32x1x1]
%onnx::Conv_783[FLOAT, 64x32x1x1]
%onnx::Conv_786[FLOAT, 64x1x5x5]
%onnx::Conv_789[FLOAT, 64x32x1x1]
%onnx::Conv_792[FLOAT, 64x64x1x1]
%onnx::Conv_795[FLOAT, 64x1x3x3]
%onnx::Conv_798[FLOAT, 112x64x1x1]
%onnx::Conv_799[FLOAT, 112]
%onnx::Conv_801[FLOAT, 112x112x1x1]
%onnx::Conv_804[FLOAT, 112x1x3x3]
%onnx::Conv_807[FLOAT, 112x112x1x1]
%onnx::Conv_810[FLOAT, 336x112x1x1]
%onnx::Conv_811[FLOAT, 336]
%onnx::Conv_813[FLOAT, 336x1x3x3]
%onnx::Conv_816[FLOAT, 112x336x1x1]
%onnx::Conv_819[FLOAT, 112x56x1x1]
%onnx::Conv_822[FLOAT, 112x1x3x3]
%onnx::Conv_825[FLOAT, 112x56x1x1]
%onnx::Conv_828[FLOAT, 336x112x1x1]
%onnx::Conv_831[FLOAT, 336x1x5x5]
%onnx::Conv_834[FLOAT, 184x336x1x1]
%onnx::Conv_835[FLOAT, 184]
%onnx::Conv_837[FLOAT, 184x184x1x1]
%onnx::Conv_840[FLOAT, 184x1x5x5]
%onnx::Conv_843[FLOAT, 184x184x1x1]
%onnx::Conv_846[FLOAT, 1104x184x1x1]
%onnx::Conv_847[FLOAT, 1104]
%onnx::Conv_849[FLOAT, 1104x1x3x3]
%onnx::Conv_852[FLOAT, 184x1104x1x1]
%onnx::Conv_855[FLOAT, 184x92x1x1]
%onnx::Conv_858[FLOAT, 184x1x5x5]
%onnx::Conv_861[FLOAT, 184x92x1x1]
%onnx::Conv_864[FLOAT, 184x92x1x1]
%onnx::Conv_867[FLOAT, 184x1x3x3]
%onnx::Conv_870[FLOAT, 352x92x1x1]
%onnx::Conv_871[FLOAT, 352]
%onnx::Conv_873[FLOAT, 1504x352x1x1]
%onnx::Conv_874[FLOAT, 1504]
) {
%onnx::Conv_868 = Identity(%onnx::Conv_835)
%onnx::Conv_865 = Identity(%onnx::Conv_835)
%onnx::Conv_862 = Identity(%onnx::Conv_835)
%onnx::Conv_859 = Identity(%onnx::Conv_835)
%onnx::Conv_856 = Identity(%onnx::Conv_835)
%onnx::Conv_853 = Identity(%onnx::Conv_835)
%onnx::Conv_850 = Identity(%onnx::Conv_847)
%onnx::Conv_844 = Identity(%onnx::Conv_835)
%onnx::Conv_841 = Identity(%onnx::Conv_835)
%onnx::Conv_838 = Identity(%onnx::Conv_835)
%onnx::Conv_832 = Identity(%onnx::Conv_811)
%onnx::Conv_829 = Identity(%onnx::Conv_811)
%onnx::Conv_826 = Identity(%onnx::Conv_799)
%onnx::Conv_823 = Identity(%onnx::Conv_799)
%onnx::Conv_820 = Identity(%onnx::Conv_799)
%onnx::Conv_817 = Identity(%onnx::Conv_799)
%onnx::Conv_814 = Identity(%onnx::Conv_811)
%onnx::Conv_808 = Identity(%onnx::Conv_799)
%onnx::Conv_805 = Identity(%onnx::Conv_799)
%onnx::Conv_802 = Identity(%onnx::Conv_799)
%onnx::Conv_796 = Identity(%onnx::Conv_772)
%onnx::Conv_793 = Identity(%onnx::Conv_772)
%onnx::Conv_790 = Identity(%onnx::Conv_772)
%onnx::Conv_787 = Identity(%onnx::Conv_772)
%onnx::Conv_784 = Identity(%onnx::Conv_772)
%onnx::Conv_781 = Identity(%onnx::Conv_772)
%onnx::Conv_778 = Identity(%onnx::Conv_772)
%onnx::Conv_775 = Identity(%onnx::Conv_772)
%onnx::Conv_769 = Identity(%onnx::Conv_742)
%onnx::Conv_766 = Identity(%onnx::Conv_742)
%onnx::Conv_763 = Identity(%onnx::Conv_742)
%onnx::Conv_760 = Identity(%onnx::Conv_742)
%onnx::Conv_757 = Identity(%onnx::Conv_742)
%onnx::Conv_754 = Identity(%onnx::Conv_742)
%onnx::Conv_751 = Identity(%onnx::Conv_742)
%onnx::Conv_748 = Identity(%onnx::Conv_745)
%onnx::Conv_739 = Identity(%onnx::Conv_736)
%onnx::Conv_733 = Identity(%onnx::Conv_706)
%onnx::Conv_730 = Identity(%onnx::Conv_706)
%onnx::Conv_727 = Identity(%onnx::Conv_706)
%onnx::Conv_724 = Identity(%onnx::Conv_706)
%onnx::Conv_721 = Identity(%onnx::Conv_706)
%onnx::Conv_718 = Identity(%onnx::Conv_706)
%onnx::Conv_715 = Identity(%onnx::Conv_706)
%onnx::Conv_712 = Identity(%onnx::Conv_709)
%onnx::Conv_703 = Identity(%onnx::Conv_700)
%onnx::Conv_697 = Identity(%onnx::Conv_688)
%onnx::Conv_694 = Identity(%onnx::Conv_688)
%onnx::Conv_691 = Identity(%onnx::Conv_688)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_687, %onnx::Conv_688)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.8/Add_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.9/relu/Relu_output_0 = Relu(%/cells.9/conv/Conv_output_0)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/relu/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/relu/Relu_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_858, %onnx::Conv_859)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_861, %onnx::Conv_862)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_864, %onnx::Conv_865)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_867, %onnx::Conv_868)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_870, %onnx::Conv_871)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_873, %onnx::Conv_874)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%685 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %685
}
|
val_accuracy
| 0
| 59,133,568
| 1,574,244
|
{'zcp_synflow': 74.83963201232892, 'zcp_zen': 64.88760375976562, 'zcp_epe_nas': 20.82337555629139, 'zcp_fisher': 0.13837577402591705, 'zcp_flops': 59133568.0, 'zcp_grad_norm': 25.489967346191406, 'zcp_grasp': -0.026889801025390625, 'zcp_jacov': -16.058455534466603, 'zcp_l2_norm': 568.849609375, 'zcp_nwot': 211.84476541999672, 'zcp_params': 1574244.0, 'zcp_plain': 0.007406145799905062, 'zcp_snip': 41.331298828125, 'lat_1080ti_1': 0.5827338578260135, 'lat_1080ti_32': 0.5632677650929478, 'lat_1080ti_64': 0.5209994362566027, 'lat_2080ti_1': 0.6979933356015501, 'lat_2080ti_32': 0.6226145293440636, 'lat_2080ti_64': 0.53491598977832, 'lat_essential_ph_1': 0.24528301886792453, 'lat_eyeriss': 0.3494764637925659, 'lat_fpga': 0.30122997456931105, 'lat_gold_6226': 0.20515669737838005, 'lat_gold_6240': 0.4632044402043129, 'lat_pixel2': 0.2391304347826087, 'lat_pixel3': 0.35244960162917255, 'lat_raspi4': 0.35333415430591925, 'lat_samsung_a50': 0.15789473684210525, 'lat_samsung_s7': 0.08661417322834646, 'lat_silver_4114': 0.4168437664044151, 'lat_silver_4210r': 0.4789523604897412, 'lat_titan_rtx_1': 0.6556293762146348, 'lat_titan_rtx_32': 0.5986679431244062, 'lat_titan_rtx_64': 0.5643269222720299, 'lat_titanx_1': 0.34280713301840293, 'lat_titanx_32': 0.5892883380349266, 'lat_titanx_64': 0.46972323497104074, 'lat_titanxp_1': 0.594606149990934, 'lat_titanxp_32': 0.6161645352811251, 'lat_titanxp_64': 0.5170762965602622}
| |
FBNet_3655
|
FBNet
|
3655
|
3655
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_578[FLOAT, 16x3x3x3]
%onnx::Conv_579[FLOAT, 16]
%onnx::Conv_581[FLOAT, 16x16x1x1]
%onnx::Conv_584[FLOAT, 16x1x3x3]
%onnx::Conv_587[FLOAT, 24x16x1x1]
%onnx::Conv_588[FLOAT, 24]
%onnx::Conv_590[FLOAT, 144x24x1x1]
%onnx::Conv_591[FLOAT, 144]
%onnx::Conv_593[FLOAT, 144x1x5x5]
%onnx::Conv_596[FLOAT, 24x144x1x1]
%onnx::Conv_599[FLOAT, 24x24x1x1]
%onnx::Conv_602[FLOAT, 24x1x5x5]
%onnx::Conv_605[FLOAT, 24x24x1x1]
%onnx::Conv_608[FLOAT, 24x24x1x1]
%onnx::Conv_611[FLOAT, 24x1x5x5]
%onnx::Conv_614[FLOAT, 24x24x1x1]
%onnx::Conv_617[FLOAT, 32x24x1x1]
%onnx::Conv_618[FLOAT, 32]
%onnx::Conv_620[FLOAT, 32x16x1x1]
%onnx::Conv_623[FLOAT, 32x1x3x3]
%onnx::Conv_626[FLOAT, 32x16x1x1]
%onnx::Conv_629[FLOAT, 96x32x1x1]
%onnx::Conv_630[FLOAT, 96]
%onnx::Conv_632[FLOAT, 96x1x5x5]
%onnx::Conv_635[FLOAT, 64x96x1x1]
%onnx::Conv_636[FLOAT, 64]
%onnx::Conv_638[FLOAT, 64x64x1x1]
%onnx::Conv_641[FLOAT, 64x1x5x5]
%onnx::Conv_644[FLOAT, 64x64x1x1]
%onnx::Conv_647[FLOAT, 64x64x1x1]
%onnx::Conv_650[FLOAT, 64x1x3x3]
%onnx::Conv_653[FLOAT, 64x64x1x1]
%onnx::Conv_656[FLOAT, 192x64x1x1]
%onnx::Conv_657[FLOAT, 192]
%onnx::Conv_659[FLOAT, 192x1x5x5]
%onnx::Conv_662[FLOAT, 64x192x1x1]
%onnx::Conv_665[FLOAT, 64x64x1x1]
%onnx::Conv_668[FLOAT, 64x1x3x3]
%onnx::Conv_671[FLOAT, 112x64x1x1]
%onnx::Conv_672[FLOAT, 112]
%onnx::Conv_674[FLOAT, 112x112x1x1]
%onnx::Conv_677[FLOAT, 112x1x3x3]
%onnx::Conv_680[FLOAT, 112x112x1x1]
%onnx::Conv_683[FLOAT, 336x112x1x1]
%onnx::Conv_684[FLOAT, 336]
%onnx::Conv_686[FLOAT, 336x1x5x5]
%onnx::Conv_689[FLOAT, 112x336x1x1]
%onnx::Conv_692[FLOAT, 112x112x1x1]
%onnx::Conv_695[FLOAT, 112x1x3x3]
%onnx::Conv_698[FLOAT, 112x112x1x1]
%onnx::Conv_701[FLOAT, 112x56x1x1]
%onnx::Conv_704[FLOAT, 112x1x3x3]
%onnx::Conv_707[FLOAT, 184x56x1x1]
%onnx::Conv_708[FLOAT, 184]
%onnx::Conv_710[FLOAT, 184x92x1x1]
%onnx::Conv_713[FLOAT, 184x1x3x3]
%onnx::Conv_716[FLOAT, 184x92x1x1]
%onnx::Conv_719[FLOAT, 552x184x1x1]
%onnx::Conv_720[FLOAT, 552]
%onnx::Conv_722[FLOAT, 552x1x5x5]
%onnx::Conv_725[FLOAT, 184x552x1x1]
%onnx::Conv_728[FLOAT, 552x184x1x1]
%onnx::Conv_731[FLOAT, 552x1x5x5]
%onnx::Conv_734[FLOAT, 184x552x1x1]
%onnx::Conv_737[FLOAT, 552x184x1x1]
%onnx::Conv_740[FLOAT, 552x1x5x5]
%onnx::Conv_743[FLOAT, 352x552x1x1]
%onnx::Conv_744[FLOAT, 352]
%onnx::Conv_746[FLOAT, 1504x352x1x1]
%onnx::Conv_747[FLOAT, 1504]
) {
%onnx::Conv_741 = Identity(%onnx::Conv_720)
%onnx::Conv_738 = Identity(%onnx::Conv_720)
%onnx::Conv_735 = Identity(%onnx::Conv_708)
%onnx::Conv_732 = Identity(%onnx::Conv_720)
%onnx::Conv_729 = Identity(%onnx::Conv_720)
%onnx::Conv_726 = Identity(%onnx::Conv_708)
%onnx::Conv_723 = Identity(%onnx::Conv_720)
%onnx::Conv_717 = Identity(%onnx::Conv_708)
%onnx::Conv_714 = Identity(%onnx::Conv_708)
%onnx::Conv_711 = Identity(%onnx::Conv_708)
%onnx::Conv_705 = Identity(%onnx::Conv_672)
%onnx::Conv_702 = Identity(%onnx::Conv_672)
%onnx::Conv_699 = Identity(%onnx::Conv_672)
%onnx::Conv_696 = Identity(%onnx::Conv_672)
%onnx::Conv_693 = Identity(%onnx::Conv_672)
%onnx::Conv_690 = Identity(%onnx::Conv_672)
%onnx::Conv_687 = Identity(%onnx::Conv_684)
%onnx::Conv_681 = Identity(%onnx::Conv_672)
%onnx::Conv_678 = Identity(%onnx::Conv_672)
%onnx::Conv_675 = Identity(%onnx::Conv_672)
%onnx::Conv_669 = Identity(%onnx::Conv_636)
%onnx::Conv_666 = Identity(%onnx::Conv_636)
%onnx::Conv_663 = Identity(%onnx::Conv_636)
%onnx::Conv_660 = Identity(%onnx::Conv_657)
%onnx::Conv_654 = Identity(%onnx::Conv_636)
%onnx::Conv_651 = Identity(%onnx::Conv_636)
%onnx::Conv_648 = Identity(%onnx::Conv_636)
%onnx::Conv_645 = Identity(%onnx::Conv_636)
%onnx::Conv_642 = Identity(%onnx::Conv_636)
%onnx::Conv_639 = Identity(%onnx::Conv_636)
%onnx::Conv_633 = Identity(%onnx::Conv_630)
%onnx::Conv_627 = Identity(%onnx::Conv_618)
%onnx::Conv_624 = Identity(%onnx::Conv_618)
%onnx::Conv_621 = Identity(%onnx::Conv_618)
%onnx::Conv_615 = Identity(%onnx::Conv_588)
%onnx::Conv_612 = Identity(%onnx::Conv_588)
%onnx::Conv_609 = Identity(%onnx::Conv_588)
%onnx::Conv_606 = Identity(%onnx::Conv_588)
%onnx::Conv_603 = Identity(%onnx::Conv_588)
%onnx::Conv_600 = Identity(%onnx::Conv_588)
%onnx::Conv_597 = Identity(%onnx::Conv_588)
%onnx::Conv_594 = Identity(%onnx::Conv_591)
%onnx::Conv_585 = Identity(%onnx::Conv_579)
%onnx::Conv_582 = Identity(%onnx::Conv_579)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_578, %onnx::Conv_579)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_581, %onnx::Conv_582)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_584, %onnx::Conv_585)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_587, %onnx::Conv_588)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_590, %onnx::Conv_591)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_593, %onnx::Conv_594)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_596, %onnx::Conv_597)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_599, %onnx::Conv_600)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_602, %onnx::Conv_603)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_605, %onnx::Conv_606)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_608, %onnx::Conv_609)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_611, %onnx::Conv_612)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_614, %onnx::Conv_615)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.4/Add_output_0, %onnx::Conv_617, %onnx::Conv_618)
%/cells.5/relu/Relu_output_0 = Relu(%/cells.5/conv/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/relu/Relu_output_0, %onnx::Conv_620, %onnx::Conv_621)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_623, %onnx::Conv_624)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.5/relu/Relu_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_output_0 = Reshape(%/cells.18/nl/Relu_output_0, %/cells.18/shuffle/Constant_output_0)
%/cells.18/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.18/shuffle/Reshape_output_0)
%/cells.18/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_1_output_0 = Reshape(%/cells.18/shuffle/Transpose_output_0, %/cells.18/shuffle/Constant_1_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/shuffle/Reshape_1_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%576 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %576
}
|
val_accuracy
| 0
| 53,802,368
| 1,720,980
|
{'zcp_synflow': 74.25512795431806, 'zcp_zen': 63.82991027832031, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.06015850231051445, 'zcp_flops': 53802368.0, 'zcp_grad_norm': 16.36539077758789, 'zcp_grasp': -0.02789020538330078, 'zcp_jacov': -16.066799580986373, 'zcp_l2_norm': 561.010986328125, 'zcp_nwot': 205.58361846157376, 'zcp_params': 1720980.0, 'zcp_plain': 0.003920369781553745, 'zcp_snip': 29.871456146240234, 'lat_1080ti_1': 0.4847826981263117, 'lat_1080ti_32': 0.3627795053487077, 'lat_1080ti_64': 0.3014357107548113, 'lat_2080ti_1': 0.4220851766768961, 'lat_2080ti_32': 0.4261556327660164, 'lat_2080ti_64': 0.31554966777310356, 'lat_essential_ph_1': 0.18867924528301888, 'lat_eyeriss': 0.2603836209267782, 'lat_fpga': 0.27786738439921227, 'lat_gold_6226': 0.22528800699299542, 'lat_gold_6240': 0.33103207179907373, 'lat_pixel2': 0.17391304347826086, 'lat_pixel3': 0.2967953989120169, 'lat_raspi4': 0.3658973597153731, 'lat_samsung_a50': 0.10526315789473684, 'lat_samsung_s7': 0.09448818897637795, 'lat_silver_4114': 0.34298147044127886, 'lat_silver_4210r': 0.31335951766299713, 'lat_titan_rtx_1': 0.3901024392697529, 'lat_titan_rtx_32': 0.4098419626989134, 'lat_titan_rtx_64': 0.33243722460099734, 'lat_titanx_1': 0.2054425533784904, 'lat_titanx_32': 0.36272850294544934, 'lat_titanx_64': 0.29696227721603796, 'lat_titanxp_1': 0.36432703185107745, 'lat_titanxp_32': 0.39622633732396606, 'lat_titanxp_64': 0.31881248497330417}
| |
FBNet_837
|
FBNet
|
837
|
837
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_624[FLOAT, 16x3x3x3]
%onnx::Conv_625[FLOAT, 16]
%onnx::Conv_627[FLOAT, 16x8x1x1]
%onnx::Conv_630[FLOAT, 16x1x5x5]
%onnx::Conv_633[FLOAT, 24x8x1x1]
%onnx::Conv_634[FLOAT, 24]
%onnx::Conv_636[FLOAT, 144x24x1x1]
%onnx::Conv_637[FLOAT, 144]
%onnx::Conv_639[FLOAT, 144x1x5x5]
%onnx::Conv_642[FLOAT, 24x144x1x1]
%onnx::Conv_645[FLOAT, 72x24x1x1]
%onnx::Conv_646[FLOAT, 72]
%onnx::Conv_648[FLOAT, 72x1x3x3]
%onnx::Conv_651[FLOAT, 24x72x1x1]
%onnx::Conv_654[FLOAT, 72x24x1x1]
%onnx::Conv_657[FLOAT, 72x1x5x5]
%onnx::Conv_660[FLOAT, 24x72x1x1]
%onnx::Conv_663[FLOAT, 144x24x1x1]
%onnx::Conv_666[FLOAT, 144x1x5x5]
%onnx::Conv_669[FLOAT, 32x144x1x1]
%onnx::Conv_670[FLOAT, 32]
%onnx::Conv_672[FLOAT, 96x32x1x1]
%onnx::Conv_673[FLOAT, 96]
%onnx::Conv_675[FLOAT, 96x1x3x3]
%onnx::Conv_678[FLOAT, 32x96x1x1]
%onnx::Conv_681[FLOAT, 32x16x1x1]
%onnx::Conv_684[FLOAT, 32x1x5x5]
%onnx::Conv_687[FLOAT, 32x16x1x1]
%onnx::Conv_690[FLOAT, 32x32x1x1]
%onnx::Conv_693[FLOAT, 32x1x5x5]
%onnx::Conv_696[FLOAT, 32x32x1x1]
%onnx::Conv_699[FLOAT, 96x32x1x1]
%onnx::Conv_702[FLOAT, 96x1x3x3]
%onnx::Conv_705[FLOAT, 64x96x1x1]
%onnx::Conv_706[FLOAT, 64]
%onnx::Conv_708[FLOAT, 192x64x1x1]
%onnx::Conv_709[FLOAT, 192]
%onnx::Conv_711[FLOAT, 192x1x3x3]
%onnx::Conv_714[FLOAT, 64x192x1x1]
%onnx::Conv_717[FLOAT, 64x64x1x1]
%onnx::Conv_720[FLOAT, 64x1x3x3]
%onnx::Conv_723[FLOAT, 64x64x1x1]
%onnx::Conv_726[FLOAT, 192x64x1x1]
%onnx::Conv_729[FLOAT, 192x1x5x5]
%onnx::Conv_732[FLOAT, 112x192x1x1]
%onnx::Conv_733[FLOAT, 112]
%onnx::Conv_735[FLOAT, 336x112x1x1]
%onnx::Conv_736[FLOAT, 336]
%onnx::Conv_738[FLOAT, 336x1x3x3]
%onnx::Conv_741[FLOAT, 112x336x1x1]
%onnx::Conv_744[FLOAT, 112x56x1x1]
%onnx::Conv_747[FLOAT, 112x1x3x3]
%onnx::Conv_750[FLOAT, 112x56x1x1]
%onnx::Conv_753[FLOAT, 672x112x1x1]
%onnx::Conv_754[FLOAT, 672]
%onnx::Conv_756[FLOAT, 672x1x5x5]
%onnx::Conv_759[FLOAT, 112x672x1x1]
%onnx::Conv_762[FLOAT, 112x112x1x1]
%onnx::Conv_765[FLOAT, 112x1x5x5]
%onnx::Conv_768[FLOAT, 184x112x1x1]
%onnx::Conv_769[FLOAT, 184]
%onnx::Conv_771[FLOAT, 184x92x1x1]
%onnx::Conv_774[FLOAT, 184x1x5x5]
%onnx::Conv_777[FLOAT, 184x92x1x1]
%onnx::Conv_780[FLOAT, 552x184x1x1]
%onnx::Conv_781[FLOAT, 552]
%onnx::Conv_783[FLOAT, 552x1x3x3]
%onnx::Conv_786[FLOAT, 184x552x1x1]
%onnx::Conv_789[FLOAT, 1104x184x1x1]
%onnx::Conv_790[FLOAT, 1104]
%onnx::Conv_792[FLOAT, 1104x1x3x3]
%onnx::Conv_795[FLOAT, 184x1104x1x1]
%onnx::Conv_798[FLOAT, 352x184x1x1]
%onnx::Conv_799[FLOAT, 352]
%onnx::Conv_801[FLOAT, 1504x352x1x1]
%onnx::Conv_802[FLOAT, 1504]
) {
%onnx::Conv_796 = Identity(%onnx::Conv_769)
%onnx::Conv_793 = Identity(%onnx::Conv_790)
%onnx::Conv_787 = Identity(%onnx::Conv_769)
%onnx::Conv_784 = Identity(%onnx::Conv_781)
%onnx::Conv_778 = Identity(%onnx::Conv_769)
%onnx::Conv_775 = Identity(%onnx::Conv_769)
%onnx::Conv_772 = Identity(%onnx::Conv_769)
%onnx::Conv_766 = Identity(%onnx::Conv_733)
%onnx::Conv_763 = Identity(%onnx::Conv_733)
%onnx::Conv_760 = Identity(%onnx::Conv_733)
%onnx::Conv_757 = Identity(%onnx::Conv_754)
%onnx::Conv_751 = Identity(%onnx::Conv_733)
%onnx::Conv_748 = Identity(%onnx::Conv_733)
%onnx::Conv_745 = Identity(%onnx::Conv_733)
%onnx::Conv_742 = Identity(%onnx::Conv_733)
%onnx::Conv_739 = Identity(%onnx::Conv_736)
%onnx::Conv_730 = Identity(%onnx::Conv_709)
%onnx::Conv_727 = Identity(%onnx::Conv_709)
%onnx::Conv_724 = Identity(%onnx::Conv_706)
%onnx::Conv_721 = Identity(%onnx::Conv_706)
%onnx::Conv_718 = Identity(%onnx::Conv_706)
%onnx::Conv_715 = Identity(%onnx::Conv_706)
%onnx::Conv_712 = Identity(%onnx::Conv_709)
%onnx::Conv_703 = Identity(%onnx::Conv_673)
%onnx::Conv_700 = Identity(%onnx::Conv_673)
%onnx::Conv_697 = Identity(%onnx::Conv_670)
%onnx::Conv_694 = Identity(%onnx::Conv_670)
%onnx::Conv_691 = Identity(%onnx::Conv_670)
%onnx::Conv_688 = Identity(%onnx::Conv_670)
%onnx::Conv_685 = Identity(%onnx::Conv_670)
%onnx::Conv_682 = Identity(%onnx::Conv_670)
%onnx::Conv_679 = Identity(%onnx::Conv_670)
%onnx::Conv_676 = Identity(%onnx::Conv_673)
%onnx::Conv_667 = Identity(%onnx::Conv_637)
%onnx::Conv_664 = Identity(%onnx::Conv_637)
%onnx::Conv_661 = Identity(%onnx::Conv_634)
%onnx::Conv_658 = Identity(%onnx::Conv_646)
%onnx::Conv_655 = Identity(%onnx::Conv_646)
%onnx::Conv_652 = Identity(%onnx::Conv_634)
%onnx::Conv_649 = Identity(%onnx::Conv_646)
%onnx::Conv_643 = Identity(%onnx::Conv_634)
%onnx::Conv_640 = Identity(%onnx::Conv_637)
%onnx::Conv_631 = Identity(%onnx::Conv_625)
%onnx::Conv_628 = Identity(%onnx::Conv_625)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_624, %onnx::Conv_625)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_627, %onnx::Conv_628)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_630, %onnx::Conv_631)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_output_0 = Reshape(%/cells.18/nl/Relu_output_0, %/cells.18/shuffle/Constant_output_0)
%/cells.18/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.18/shuffle/Reshape_output_0)
%/cells.18/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_1_output_0 = Reshape(%/cells.18/shuffle/Transpose_output_0, %/cells.18/shuffle/Constant_1_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/shuffle/Reshape_1_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.21/relu/Relu_output_0 = Relu(%/cells.21/conv/Conv_output_0)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/relu/Relu_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%622 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %622
}
|
val_accuracy
| 0
| 77,566,848
| 1,854,516
|
{'zcp_synflow': 76.76784476690507, 'zcp_zen': 67.43156433105469, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.06126142665743828, 'zcp_flops': 77566848.0, 'zcp_grad_norm': 18.889705657958984, 'zcp_grasp': -0.05791664123535156, 'zcp_jacov': -16.059734823232528, 'zcp_l2_norm': 614.312255859375, 'zcp_nwot': 214.9205287944121, 'zcp_params': 1854516.0, 'zcp_plain': -0.0020953312050551176, 'zcp_snip': 35.31832504272461, 'lat_1080ti_1': 0.5746228961592624, 'lat_1080ti_32': 0.5728539399802551, 'lat_1080ti_64': 0.5992326012999676, 'lat_2080ti_1': 0.5356370839746546, 'lat_2080ti_32': 0.5550781805158679, 'lat_2080ti_64': 0.5789991692117288, 'lat_essential_ph_1': 0.3584905660377358, 'lat_eyeriss': 0.5255441659178848, 'lat_fpga': 0.5307592594425037, 'lat_gold_6226': 0.5095658683057036, 'lat_gold_6240': 0.43116432870560106, 'lat_pixel2': 0.5652173913043478, 'lat_pixel3': 0.5710680900980042, 'lat_raspi4': 0.5703272701470201, 'lat_samsung_a50': 0.22105263157894736, 'lat_samsung_s7': 0.18110236220472442, 'lat_silver_4114': 0.4730419620211324, 'lat_silver_4210r': 0.4102488486873299, 'lat_titan_rtx_1': 0.5056506435058833, 'lat_titan_rtx_32': 0.5070813315495691, 'lat_titan_rtx_64': 0.5852936805770562, 'lat_titanx_1': 0.265859552442063, 'lat_titanx_32': 0.6000328350743256, 'lat_titanx_64': 0.5805030552147712, 'lat_titanxp_1': 0.46755474105496114, 'lat_titanxp_32': 0.5843048138048642, 'lat_titanxp_64': 0.5955622600414247}
| |
FBNet_3113
|
FBNet
|
3113
|
3113
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_569[FLOAT, 16x3x3x3]
%onnx::Conv_570[FLOAT, 16]
%onnx::Conv_572[FLOAT, 96x16x1x1]
%onnx::Conv_573[FLOAT, 96]
%onnx::Conv_575[FLOAT, 96x1x5x5]
%onnx::Conv_578[FLOAT, 16x96x1x1]
%onnx::Conv_581[FLOAT, 24x16x1x1]
%onnx::Conv_582[FLOAT, 24]
%onnx::Conv_584[FLOAT, 24x12x1x1]
%onnx::Conv_587[FLOAT, 24x1x3x3]
%onnx::Conv_590[FLOAT, 24x12x1x1]
%onnx::Conv_593[FLOAT, 72x24x1x1]
%onnx::Conv_594[FLOAT, 72]
%onnx::Conv_596[FLOAT, 72x1x5x5]
%onnx::Conv_599[FLOAT, 24x72x1x1]
%onnx::Conv_602[FLOAT, 72x24x1x1]
%onnx::Conv_605[FLOAT, 72x1x3x3]
%onnx::Conv_608[FLOAT, 24x72x1x1]
%onnx::Conv_611[FLOAT, 72x24x1x1]
%onnx::Conv_614[FLOAT, 72x1x5x5]
%onnx::Conv_617[FLOAT, 32x72x1x1]
%onnx::Conv_618[FLOAT, 32]
%onnx::Conv_620[FLOAT, 192x32x1x1]
%onnx::Conv_621[FLOAT, 192]
%onnx::Conv_623[FLOAT, 192x1x3x3]
%onnx::Conv_626[FLOAT, 32x192x1x1]
%onnx::Conv_629[FLOAT, 32x32x1x1]
%onnx::Conv_632[FLOAT, 32x1x3x3]
%onnx::Conv_635[FLOAT, 32x32x1x1]
%onnx::Conv_638[FLOAT, 32x32x1x1]
%onnx::Conv_641[FLOAT, 32x1x3x3]
%onnx::Conv_644[FLOAT, 64x32x1x1]
%onnx::Conv_645[FLOAT, 64]
%onnx::Conv_647[FLOAT, 64x32x1x1]
%onnx::Conv_650[FLOAT, 64x1x3x3]
%onnx::Conv_653[FLOAT, 64x32x1x1]
%onnx::Conv_656[FLOAT, 192x64x1x1]
%onnx::Conv_659[FLOAT, 192x1x5x5]
%onnx::Conv_662[FLOAT, 64x192x1x1]
%onnx::Conv_665[FLOAT, 64x64x1x1]
%onnx::Conv_668[FLOAT, 64x1x5x5]
%onnx::Conv_671[FLOAT, 112x64x1x1]
%onnx::Conv_672[FLOAT, 112]
%onnx::Conv_674[FLOAT, 672x112x1x1]
%onnx::Conv_675[FLOAT, 672]
%onnx::Conv_677[FLOAT, 672x1x3x3]
%onnx::Conv_680[FLOAT, 112x672x1x1]
%onnx::Conv_683[FLOAT, 336x112x1x1]
%onnx::Conv_684[FLOAT, 336]
%onnx::Conv_686[FLOAT, 336x1x5x5]
%onnx::Conv_689[FLOAT, 112x336x1x1]
%onnx::Conv_692[FLOAT, 112x112x1x1]
%onnx::Conv_695[FLOAT, 112x1x3x3]
%onnx::Conv_698[FLOAT, 112x112x1x1]
%onnx::Conv_701[FLOAT, 184x112x1x1]
%onnx::Conv_702[FLOAT, 184]
%onnx::Conv_704[FLOAT, 1104x184x1x1]
%onnx::Conv_705[FLOAT, 1104]
%onnx::Conv_707[FLOAT, 1104x1x3x3]
%onnx::Conv_710[FLOAT, 184x1104x1x1]
%onnx::Conv_713[FLOAT, 184x184x1x1]
%onnx::Conv_716[FLOAT, 184x1x5x5]
%onnx::Conv_719[FLOAT, 184x184x1x1]
%onnx::Conv_722[FLOAT, 1104x184x1x1]
%onnx::Conv_725[FLOAT, 1104x1x5x5]
%onnx::Conv_728[FLOAT, 184x1104x1x1]
%onnx::Conv_731[FLOAT, 1104x184x1x1]
%onnx::Conv_734[FLOAT, 1104x1x5x5]
%onnx::Conv_737[FLOAT, 352x1104x1x1]
%onnx::Conv_738[FLOAT, 352]
%onnx::Conv_740[FLOAT, 1504x352x1x1]
%onnx::Conv_741[FLOAT, 1504]
) {
%onnx::Conv_735 = Identity(%onnx::Conv_705)
%onnx::Conv_732 = Identity(%onnx::Conv_705)
%onnx::Conv_729 = Identity(%onnx::Conv_702)
%onnx::Conv_726 = Identity(%onnx::Conv_705)
%onnx::Conv_723 = Identity(%onnx::Conv_705)
%onnx::Conv_720 = Identity(%onnx::Conv_702)
%onnx::Conv_717 = Identity(%onnx::Conv_702)
%onnx::Conv_714 = Identity(%onnx::Conv_702)
%onnx::Conv_711 = Identity(%onnx::Conv_702)
%onnx::Conv_708 = Identity(%onnx::Conv_705)
%onnx::Conv_699 = Identity(%onnx::Conv_672)
%onnx::Conv_696 = Identity(%onnx::Conv_672)
%onnx::Conv_693 = Identity(%onnx::Conv_672)
%onnx::Conv_690 = Identity(%onnx::Conv_672)
%onnx::Conv_687 = Identity(%onnx::Conv_684)
%onnx::Conv_681 = Identity(%onnx::Conv_672)
%onnx::Conv_678 = Identity(%onnx::Conv_675)
%onnx::Conv_669 = Identity(%onnx::Conv_645)
%onnx::Conv_666 = Identity(%onnx::Conv_645)
%onnx::Conv_663 = Identity(%onnx::Conv_645)
%onnx::Conv_660 = Identity(%onnx::Conv_621)
%onnx::Conv_657 = Identity(%onnx::Conv_621)
%onnx::Conv_654 = Identity(%onnx::Conv_645)
%onnx::Conv_651 = Identity(%onnx::Conv_645)
%onnx::Conv_648 = Identity(%onnx::Conv_645)
%onnx::Conv_642 = Identity(%onnx::Conv_618)
%onnx::Conv_639 = Identity(%onnx::Conv_618)
%onnx::Conv_636 = Identity(%onnx::Conv_618)
%onnx::Conv_633 = Identity(%onnx::Conv_618)
%onnx::Conv_630 = Identity(%onnx::Conv_618)
%onnx::Conv_627 = Identity(%onnx::Conv_618)
%onnx::Conv_624 = Identity(%onnx::Conv_621)
%onnx::Conv_615 = Identity(%onnx::Conv_594)
%onnx::Conv_612 = Identity(%onnx::Conv_594)
%onnx::Conv_609 = Identity(%onnx::Conv_582)
%onnx::Conv_606 = Identity(%onnx::Conv_594)
%onnx::Conv_603 = Identity(%onnx::Conv_594)
%onnx::Conv_600 = Identity(%onnx::Conv_582)
%onnx::Conv_597 = Identity(%onnx::Conv_594)
%onnx::Conv_591 = Identity(%onnx::Conv_582)
%onnx::Conv_588 = Identity(%onnx::Conv_582)
%onnx::Conv_585 = Identity(%onnx::Conv_582)
%onnx::Conv_579 = Identity(%onnx::Conv_570)
%onnx::Conv_576 = Identity(%onnx::Conv_573)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_569, %onnx::Conv_570)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_572, %onnx::Conv_573)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_575, %onnx::Conv_576)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_578, %onnx::Conv_579)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_581, %onnx::Conv_582)
%/cells.1/relu/Relu_output_0 = Relu(%/cells.1/conv/Conv_output_0)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/relu/Relu_output_0, %onnx::Conv_584, %onnx::Conv_585)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_587, %onnx::Conv_588)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_590, %onnx::Conv_591)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/relu/Relu_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_593, %onnx::Conv_594)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_596, %onnx::Conv_597)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_599, %onnx::Conv_600)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_602, %onnx::Conv_603)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_605, %onnx::Conv_606)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_608, %onnx::Conv_609)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_611, %onnx::Conv_612)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_614, %onnx::Conv_615)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_617, %onnx::Conv_618)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_620, %onnx::Conv_621)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_623, %onnx::Conv_624)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%567 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %567
}
|
val_accuracy
| 0
| 81,091,072
| 2,626,508
|
{'zcp_synflow': 76.38435082650567, 'zcp_zen': 66.03121948242188, 'zcp_epe_nas': 10.04642232980952, 'zcp_fisher': 0.12261871993541718, 'zcp_flops': 81091072.0, 'zcp_grad_norm': 24.116622924804688, 'zcp_grasp': -0.14504241943359375, 'zcp_jacov': -16.04135558586115, 'zcp_l2_norm': 649.74365234375, 'zcp_nwot': 213.47371987407934, 'zcp_params': 2626508.0, 'zcp_plain': 0.005967192351818085, 'zcp_snip': 42.288578033447266, 'lat_1080ti_1': 0.443425446788664, 'lat_1080ti_32': 0.3899647312788576, 'lat_1080ti_64': 0.4318929207962135, 'lat_2080ti_1': 0.4370945073386351, 'lat_2080ti_32': 0.3986078326174457, 'lat_2080ti_64': 0.4236756680084874, 'lat_essential_ph_1': 0.660377358490566, 'lat_eyeriss': 0.594924174353144, 'lat_fpga': 0.6644930189295362, 'lat_gold_6226': 0.5078492222207444, 'lat_gold_6240': 0.56223381762148, 'lat_pixel2': 0.391304347826087, 'lat_pixel3': 0.5774318461387428, 'lat_raspi4': 0.6478452256139261, 'lat_samsung_a50': 0.24210526315789474, 'lat_samsung_s7': 0.2204724409448819, 'lat_silver_4114': 0.5668361097472171, 'lat_silver_4210r': 0.5596154730342212, 'lat_titan_rtx_1': 0.4202262120592683, 'lat_titan_rtx_32': 0.37624679274202516, 'lat_titan_rtx_64': 0.40065601357575603, 'lat_titanx_1': 0.23240318592926065, 'lat_titanx_32': 0.4013032060974374, 'lat_titanx_64': 0.4324485898673198, 'lat_titanxp_1': 0.40179244741502534, 'lat_titanxp_32': 0.40060303731498265, 'lat_titanxp_64': 0.41312916993758453}
| |
FBNet_4657
|
FBNet
|
4657
|
4657
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_680[FLOAT, 16x3x3x3]
%onnx::Conv_681[FLOAT, 16]
%onnx::Conv_683[FLOAT, 16x8x1x1]
%onnx::Conv_686[FLOAT, 16x1x3x3]
%onnx::Conv_689[FLOAT, 16x8x1x1]
%onnx::Conv_692[FLOAT, 24x16x1x1]
%onnx::Conv_693[FLOAT, 24]
%onnx::Conv_695[FLOAT, 144x24x1x1]
%onnx::Conv_696[FLOAT, 144]
%onnx::Conv_698[FLOAT, 144x1x3x3]
%onnx::Conv_701[FLOAT, 24x144x1x1]
%onnx::Conv_704[FLOAT, 72x24x1x1]
%onnx::Conv_705[FLOAT, 72]
%onnx::Conv_707[FLOAT, 72x1x5x5]
%onnx::Conv_710[FLOAT, 24x72x1x1]
%onnx::Conv_713[FLOAT, 144x24x1x1]
%onnx::Conv_716[FLOAT, 144x1x3x3]
%onnx::Conv_719[FLOAT, 24x144x1x1]
%onnx::Conv_722[FLOAT, 72x24x1x1]
%onnx::Conv_725[FLOAT, 72x1x3x3]
%onnx::Conv_728[FLOAT, 32x72x1x1]
%onnx::Conv_729[FLOAT, 32]
%onnx::Conv_731[FLOAT, 32x16x1x1]
%onnx::Conv_734[FLOAT, 32x1x5x5]
%onnx::Conv_737[FLOAT, 32x16x1x1]
%onnx::Conv_740[FLOAT, 32x32x1x1]
%onnx::Conv_743[FLOAT, 32x1x5x5]
%onnx::Conv_746[FLOAT, 32x32x1x1]
%onnx::Conv_749[FLOAT, 96x32x1x1]
%onnx::Conv_750[FLOAT, 96]
%onnx::Conv_752[FLOAT, 96x1x5x5]
%onnx::Conv_755[FLOAT, 32x96x1x1]
%onnx::Conv_758[FLOAT, 32x32x1x1]
%onnx::Conv_761[FLOAT, 32x1x3x3]
%onnx::Conv_764[FLOAT, 64x32x1x1]
%onnx::Conv_765[FLOAT, 64]
%onnx::Conv_767[FLOAT, 384x64x1x1]
%onnx::Conv_768[FLOAT, 384]
%onnx::Conv_770[FLOAT, 384x1x3x3]
%onnx::Conv_773[FLOAT, 64x384x1x1]
%onnx::Conv_776[FLOAT, 192x64x1x1]
%onnx::Conv_777[FLOAT, 192]
%onnx::Conv_779[FLOAT, 192x1x5x5]
%onnx::Conv_782[FLOAT, 64x192x1x1]
%onnx::Conv_785[FLOAT, 192x64x1x1]
%onnx::Conv_788[FLOAT, 192x1x3x3]
%onnx::Conv_791[FLOAT, 64x192x1x1]
%onnx::Conv_794[FLOAT, 64x32x1x1]
%onnx::Conv_797[FLOAT, 64x1x5x5]
%onnx::Conv_800[FLOAT, 112x32x1x1]
%onnx::Conv_801[FLOAT, 112]
%onnx::Conv_803[FLOAT, 112x56x1x1]
%onnx::Conv_806[FLOAT, 112x1x5x5]
%onnx::Conv_809[FLOAT, 112x56x1x1]
%onnx::Conv_812[FLOAT, 672x112x1x1]
%onnx::Conv_813[FLOAT, 672]
%onnx::Conv_815[FLOAT, 672x1x3x3]
%onnx::Conv_818[FLOAT, 112x672x1x1]
%onnx::Conv_821[FLOAT, 672x112x1x1]
%onnx::Conv_824[FLOAT, 672x1x3x3]
%onnx::Conv_827[FLOAT, 112x672x1x1]
%onnx::Conv_830[FLOAT, 184x112x1x1]
%onnx::Conv_831[FLOAT, 184]
%onnx::Conv_833[FLOAT, 1104x184x1x1]
%onnx::Conv_834[FLOAT, 1104]
%onnx::Conv_836[FLOAT, 1104x1x5x5]
%onnx::Conv_839[FLOAT, 184x1104x1x1]
%onnx::Conv_842[FLOAT, 184x184x1x1]
%onnx::Conv_845[FLOAT, 184x1x5x5]
%onnx::Conv_848[FLOAT, 184x184x1x1]
%onnx::Conv_851[FLOAT, 184x92x1x1]
%onnx::Conv_854[FLOAT, 184x1x5x5]
%onnx::Conv_857[FLOAT, 184x92x1x1]
%onnx::Conv_860[FLOAT, 1104x184x1x1]
%onnx::Conv_863[FLOAT, 1104x1x5x5]
%onnx::Conv_866[FLOAT, 352x1104x1x1]
%onnx::Conv_867[FLOAT, 352]
%onnx::Conv_869[FLOAT, 1504x352x1x1]
%onnx::Conv_870[FLOAT, 1504]
) {
%onnx::Conv_864 = Identity(%onnx::Conv_834)
%onnx::Conv_861 = Identity(%onnx::Conv_834)
%onnx::Conv_858 = Identity(%onnx::Conv_831)
%onnx::Conv_855 = Identity(%onnx::Conv_831)
%onnx::Conv_852 = Identity(%onnx::Conv_831)
%onnx::Conv_849 = Identity(%onnx::Conv_831)
%onnx::Conv_846 = Identity(%onnx::Conv_831)
%onnx::Conv_843 = Identity(%onnx::Conv_831)
%onnx::Conv_840 = Identity(%onnx::Conv_831)
%onnx::Conv_837 = Identity(%onnx::Conv_834)
%onnx::Conv_828 = Identity(%onnx::Conv_801)
%onnx::Conv_825 = Identity(%onnx::Conv_813)
%onnx::Conv_822 = Identity(%onnx::Conv_813)
%onnx::Conv_819 = Identity(%onnx::Conv_801)
%onnx::Conv_816 = Identity(%onnx::Conv_813)
%onnx::Conv_810 = Identity(%onnx::Conv_801)
%onnx::Conv_807 = Identity(%onnx::Conv_801)
%onnx::Conv_804 = Identity(%onnx::Conv_801)
%onnx::Conv_798 = Identity(%onnx::Conv_765)
%onnx::Conv_795 = Identity(%onnx::Conv_765)
%onnx::Conv_792 = Identity(%onnx::Conv_765)
%onnx::Conv_789 = Identity(%onnx::Conv_777)
%onnx::Conv_786 = Identity(%onnx::Conv_777)
%onnx::Conv_783 = Identity(%onnx::Conv_765)
%onnx::Conv_780 = Identity(%onnx::Conv_777)
%onnx::Conv_774 = Identity(%onnx::Conv_765)
%onnx::Conv_771 = Identity(%onnx::Conv_768)
%onnx::Conv_762 = Identity(%onnx::Conv_729)
%onnx::Conv_759 = Identity(%onnx::Conv_729)
%onnx::Conv_756 = Identity(%onnx::Conv_729)
%onnx::Conv_753 = Identity(%onnx::Conv_750)
%onnx::Conv_747 = Identity(%onnx::Conv_729)
%onnx::Conv_744 = Identity(%onnx::Conv_729)
%onnx::Conv_741 = Identity(%onnx::Conv_729)
%onnx::Conv_738 = Identity(%onnx::Conv_729)
%onnx::Conv_735 = Identity(%onnx::Conv_729)
%onnx::Conv_732 = Identity(%onnx::Conv_729)
%onnx::Conv_726 = Identity(%onnx::Conv_705)
%onnx::Conv_723 = Identity(%onnx::Conv_705)
%onnx::Conv_720 = Identity(%onnx::Conv_693)
%onnx::Conv_717 = Identity(%onnx::Conv_696)
%onnx::Conv_714 = Identity(%onnx::Conv_696)
%onnx::Conv_711 = Identity(%onnx::Conv_693)
%onnx::Conv_708 = Identity(%onnx::Conv_705)
%onnx::Conv_702 = Identity(%onnx::Conv_693)
%onnx::Conv_699 = Identity(%onnx::Conv_696)
%onnx::Conv_690 = Identity(%onnx::Conv_681)
%onnx::Conv_687 = Identity(%onnx::Conv_681)
%onnx::Conv_684 = Identity(%onnx::Conv_681)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_680, %onnx::Conv_681)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.1/relu/Relu_output_0 = Relu(%/cells.1/conv/Conv_output_0)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/relu/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/relu/Relu_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_803, %onnx::Conv_804)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_806, %onnx::Conv_807)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_809, %onnx::Conv_810)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_812, %onnx::Conv_813)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_815, %onnx::Conv_816)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_818, %onnx::Conv_819)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_821, %onnx::Conv_822)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_824, %onnx::Conv_825)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_827, %onnx::Conv_828)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_830, %onnx::Conv_831)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_833, %onnx::Conv_834)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_836, %onnx::Conv_837)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_839, %onnx::Conv_840)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_842, %onnx::Conv_843)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_845, %onnx::Conv_846)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_848, %onnx::Conv_849)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_851, %onnx::Conv_852)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_854, %onnx::Conv_855)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_857, %onnx::Conv_858)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_860, %onnx::Conv_861)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_863, %onnx::Conv_864)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_866, %onnx::Conv_867)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_869, %onnx::Conv_870)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%678 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %678
}
|
val_accuracy
| 0
| 89,635,072
| 2,381,780
|
{'zcp_synflow': 79.71715802476797, 'zcp_zen': 70.8860855102539, 'zcp_epe_nas': 14.124515759570007, 'zcp_fisher': 0.17802000045776367, 'zcp_flops': 89635072.0, 'zcp_grad_norm': 25.141616821289062, 'zcp_grasp': -0.36246490478515625, 'zcp_jacov': -16.060935652700365, 'zcp_l2_norm': 676.4277954101562, 'zcp_nwot': 217.05451323713604, 'zcp_params': 2381780.0, 'zcp_plain': -0.003351407591253519, 'zcp_snip': 49.392547607421875, 'lat_1080ti_1': 0.6454490224994694, 'lat_1080ti_32': 0.7042376253066516, 'lat_1080ti_64': 0.6339295218759599, 'lat_2080ti_1': 0.7187095541635006, 'lat_2080ti_32': 0.7362541434044386, 'lat_2080ti_64': 0.6816112385095354, 'lat_essential_ph_1': 0.4339622641509434, 'lat_eyeriss': 0.6559147672051785, 'lat_fpga': 0.7782087691348618, 'lat_gold_6226': 0.6924190445176323, 'lat_gold_6240': 0.6209933776153054, 'lat_pixel2': 0.5217391304347826, 'lat_pixel3': 0.6343263245410231, 'lat_raspi4': 0.7411738414738541, 'lat_samsung_a50': 0.28421052631578947, 'lat_samsung_s7': 0.25196850393700787, 'lat_silver_4114': 0.7035342804037238, 'lat_silver_4210r': 0.645599086520629, 'lat_titan_rtx_1': 0.6753815564134076, 'lat_titan_rtx_32': 0.6717177115822092, 'lat_titan_rtx_64': 0.6723083174160384, 'lat_titanx_1': 0.36076455487804154, 'lat_titanx_32': 0.6961608668547932, 'lat_titanx_64': 0.6194216379506425, 'lat_titanxp_1': 0.6383601683876085, 'lat_titanxp_32': 0.6974592802231929, 'lat_titanxp_64': 0.631436707122422}
| |
FBNet_4676
|
FBNet
|
4676
|
4676
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_676[FLOAT, 16x3x3x3]
%onnx::Conv_677[FLOAT, 16]
%onnx::Conv_679[FLOAT, 48x16x1x1]
%onnx::Conv_680[FLOAT, 48]
%onnx::Conv_682[FLOAT, 48x1x5x5]
%onnx::Conv_685[FLOAT, 16x48x1x1]
%onnx::Conv_688[FLOAT, 16x16x1x1]
%onnx::Conv_691[FLOAT, 16x1x3x3]
%onnx::Conv_694[FLOAT, 24x16x1x1]
%onnx::Conv_695[FLOAT, 24]
%onnx::Conv_697[FLOAT, 72x24x1x1]
%onnx::Conv_698[FLOAT, 72]
%onnx::Conv_700[FLOAT, 72x1x3x3]
%onnx::Conv_703[FLOAT, 24x72x1x1]
%onnx::Conv_706[FLOAT, 24x24x1x1]
%onnx::Conv_709[FLOAT, 24x1x3x3]
%onnx::Conv_712[FLOAT, 24x24x1x1]
%onnx::Conv_715[FLOAT, 24x24x1x1]
%onnx::Conv_718[FLOAT, 24x1x3x3]
%onnx::Conv_721[FLOAT, 32x24x1x1]
%onnx::Conv_722[FLOAT, 32]
%onnx::Conv_724[FLOAT, 32x32x1x1]
%onnx::Conv_727[FLOAT, 32x1x5x5]
%onnx::Conv_730[FLOAT, 32x32x1x1]
%onnx::Conv_733[FLOAT, 32x16x1x1]
%onnx::Conv_736[FLOAT, 32x1x3x3]
%onnx::Conv_739[FLOAT, 32x16x1x1]
%onnx::Conv_742[FLOAT, 192x32x1x1]
%onnx::Conv_743[FLOAT, 192]
%onnx::Conv_745[FLOAT, 192x1x5x5]
%onnx::Conv_748[FLOAT, 64x192x1x1]
%onnx::Conv_749[FLOAT, 64]
%onnx::Conv_751[FLOAT, 192x64x1x1]
%onnx::Conv_754[FLOAT, 192x1x3x3]
%onnx::Conv_757[FLOAT, 64x192x1x1]
%onnx::Conv_760[FLOAT, 64x64x1x1]
%onnx::Conv_763[FLOAT, 64x1x3x3]
%onnx::Conv_766[FLOAT, 64x64x1x1]
%onnx::Conv_769[FLOAT, 64x64x1x1]
%onnx::Conv_772[FLOAT, 64x1x5x5]
%onnx::Conv_775[FLOAT, 64x64x1x1]
%onnx::Conv_778[FLOAT, 384x64x1x1]
%onnx::Conv_779[FLOAT, 384]
%onnx::Conv_781[FLOAT, 384x1x5x5]
%onnx::Conv_784[FLOAT, 112x384x1x1]
%onnx::Conv_785[FLOAT, 112]
%onnx::Conv_787[FLOAT, 672x112x1x1]
%onnx::Conv_788[FLOAT, 672]
%onnx::Conv_790[FLOAT, 672x1x5x5]
%onnx::Conv_793[FLOAT, 112x672x1x1]
%onnx::Conv_796[FLOAT, 112x56x1x1]
%onnx::Conv_799[FLOAT, 112x1x5x5]
%onnx::Conv_802[FLOAT, 112x56x1x1]
%onnx::Conv_805[FLOAT, 112x56x1x1]
%onnx::Conv_808[FLOAT, 112x1x5x5]
%onnx::Conv_811[FLOAT, 112x56x1x1]
%onnx::Conv_814[FLOAT, 112x56x1x1]
%onnx::Conv_817[FLOAT, 112x1x3x3]
%onnx::Conv_820[FLOAT, 184x56x1x1]
%onnx::Conv_821[FLOAT, 184]
%onnx::Conv_823[FLOAT, 552x184x1x1]
%onnx::Conv_824[FLOAT, 552]
%onnx::Conv_826[FLOAT, 552x1x3x3]
%onnx::Conv_829[FLOAT, 184x552x1x1]
%onnx::Conv_832[FLOAT, 1104x184x1x1]
%onnx::Conv_833[FLOAT, 1104]
%onnx::Conv_835[FLOAT, 1104x1x3x3]
%onnx::Conv_838[FLOAT, 184x1104x1x1]
%onnx::Conv_841[FLOAT, 184x92x1x1]
%onnx::Conv_844[FLOAT, 184x1x5x5]
%onnx::Conv_847[FLOAT, 184x92x1x1]
%onnx::Conv_850[FLOAT, 184x92x1x1]
%onnx::Conv_853[FLOAT, 184x1x3x3]
%onnx::Conv_856[FLOAT, 352x92x1x1]
%onnx::Conv_857[FLOAT, 352]
%onnx::Conv_859[FLOAT, 1504x352x1x1]
%onnx::Conv_860[FLOAT, 1504]
) {
%onnx::Conv_854 = Identity(%onnx::Conv_821)
%onnx::Conv_851 = Identity(%onnx::Conv_821)
%onnx::Conv_848 = Identity(%onnx::Conv_821)
%onnx::Conv_845 = Identity(%onnx::Conv_821)
%onnx::Conv_842 = Identity(%onnx::Conv_821)
%onnx::Conv_839 = Identity(%onnx::Conv_821)
%onnx::Conv_836 = Identity(%onnx::Conv_833)
%onnx::Conv_830 = Identity(%onnx::Conv_821)
%onnx::Conv_827 = Identity(%onnx::Conv_824)
%onnx::Conv_818 = Identity(%onnx::Conv_785)
%onnx::Conv_815 = Identity(%onnx::Conv_785)
%onnx::Conv_812 = Identity(%onnx::Conv_785)
%onnx::Conv_809 = Identity(%onnx::Conv_785)
%onnx::Conv_806 = Identity(%onnx::Conv_785)
%onnx::Conv_803 = Identity(%onnx::Conv_785)
%onnx::Conv_800 = Identity(%onnx::Conv_785)
%onnx::Conv_797 = Identity(%onnx::Conv_785)
%onnx::Conv_794 = Identity(%onnx::Conv_785)
%onnx::Conv_791 = Identity(%onnx::Conv_788)
%onnx::Conv_782 = Identity(%onnx::Conv_779)
%onnx::Conv_776 = Identity(%onnx::Conv_749)
%onnx::Conv_773 = Identity(%onnx::Conv_749)
%onnx::Conv_770 = Identity(%onnx::Conv_749)
%onnx::Conv_767 = Identity(%onnx::Conv_749)
%onnx::Conv_764 = Identity(%onnx::Conv_749)
%onnx::Conv_761 = Identity(%onnx::Conv_749)
%onnx::Conv_758 = Identity(%onnx::Conv_749)
%onnx::Conv_755 = Identity(%onnx::Conv_743)
%onnx::Conv_752 = Identity(%onnx::Conv_743)
%onnx::Conv_746 = Identity(%onnx::Conv_743)
%onnx::Conv_740 = Identity(%onnx::Conv_722)
%onnx::Conv_737 = Identity(%onnx::Conv_722)
%onnx::Conv_734 = Identity(%onnx::Conv_722)
%onnx::Conv_731 = Identity(%onnx::Conv_722)
%onnx::Conv_728 = Identity(%onnx::Conv_722)
%onnx::Conv_725 = Identity(%onnx::Conv_722)
%onnx::Conv_719 = Identity(%onnx::Conv_695)
%onnx::Conv_716 = Identity(%onnx::Conv_695)
%onnx::Conv_713 = Identity(%onnx::Conv_695)
%onnx::Conv_710 = Identity(%onnx::Conv_695)
%onnx::Conv_707 = Identity(%onnx::Conv_695)
%onnx::Conv_704 = Identity(%onnx::Conv_695)
%onnx::Conv_701 = Identity(%onnx::Conv_698)
%onnx::Conv_692 = Identity(%onnx::Conv_677)
%onnx::Conv_689 = Identity(%onnx::Conv_677)
%onnx::Conv_686 = Identity(%onnx::Conv_677)
%onnx::Conv_683 = Identity(%onnx::Conv_680)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_676, %onnx::Conv_677)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_823, %onnx::Conv_824)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_826, %onnx::Conv_827)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_829, %onnx::Conv_830)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_832, %onnx::Conv_833)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_835, %onnx::Conv_836)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_838, %onnx::Conv_839)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_841, %onnx::Conv_842)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_844, %onnx::Conv_845)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_847, %onnx::Conv_848)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_850, %onnx::Conv_851)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_853, %onnx::Conv_854)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_856, %onnx::Conv_857)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_859, %onnx::Conv_860)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%674 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %674
}
|
val_accuracy
| 0
| 57,627,776
| 1,793,364
|
{'zcp_synflow': 74.37168167546058, 'zcp_zen': 65.51284790039062, 'zcp_epe_nas': 22.37580847168576, 'zcp_fisher': 0.12011285871267319, 'zcp_flops': 57627776.0, 'zcp_grad_norm': 22.968379974365234, 'zcp_grasp': -0.084075927734375, 'zcp_jacov': -16.057493246130424, 'zcp_l2_norm': 589.70849609375, 'zcp_nwot': 207.45454112036424, 'zcp_params': 1793364.0, 'zcp_plain': 0.00048132421215996146, 'zcp_snip': 41.891441345214844, 'lat_1080ti_1': 0.6556680184445969, 'lat_1080ti_32': 0.5082444673113871, 'lat_1080ti_64': 0.31897585478954005, 'lat_2080ti_1': 0.6502414498747382, 'lat_2080ti_32': 0.5099967931598482, 'lat_2080ti_64': 0.34640314885435025, 'lat_essential_ph_1': 0.2830188679245283, 'lat_eyeriss': 0.2892102029963508, 'lat_fpga': 0.3507752896822648, 'lat_gold_6226': 0.3099268442749088, 'lat_gold_6240': 0.436880994685756, 'lat_pixel2': 0.2391304347826087, 'lat_pixel3': 0.2829744468486933, 'lat_raspi4': 0.30164289822933626, 'lat_samsung_a50': 0.14736842105263157, 'lat_samsung_s7': 0.15748031496062992, 'lat_silver_4114': 0.46772127195164714, 'lat_silver_4210r': 0.490853653912113, 'lat_titan_rtx_1': 0.61248768826566, 'lat_titan_rtx_32': 0.5050190809658527, 'lat_titan_rtx_64': 0.38495081507520745, 'lat_titanx_1': 0.3220523495617854, 'lat_titanx_32': 0.4236549943619829, 'lat_titanx_64': 0.3091756967511221, 'lat_titanxp_1': 0.5683761557408028, 'lat_titanxp_32': 0.4606605747692884, 'lat_titanxp_64': 0.34500317253515783}
| |
FBNet_3330
|
FBNet
|
3330
|
3330
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_561[FLOAT, 16x3x3x3]
%onnx::Conv_562[FLOAT, 16]
%onnx::Conv_564[FLOAT, 16x16x1x1]
%onnx::Conv_567[FLOAT, 16x1x3x3]
%onnx::Conv_570[FLOAT, 16x16x1x1]
%onnx::Conv_573[FLOAT, 48x16x1x1]
%onnx::Conv_574[FLOAT, 48]
%onnx::Conv_576[FLOAT, 48x1x5x5]
%onnx::Conv_579[FLOAT, 24x48x1x1]
%onnx::Conv_580[FLOAT, 24]
%onnx::Conv_582[FLOAT, 24x12x1x1]
%onnx::Conv_585[FLOAT, 24x1x5x5]
%onnx::Conv_588[FLOAT, 24x12x1x1]
%onnx::Conv_591[FLOAT, 72x24x1x1]
%onnx::Conv_592[FLOAT, 72]
%onnx::Conv_594[FLOAT, 72x1x3x3]
%onnx::Conv_597[FLOAT, 24x72x1x1]
%onnx::Conv_600[FLOAT, 72x24x1x1]
%onnx::Conv_603[FLOAT, 72x1x5x5]
%onnx::Conv_606[FLOAT, 24x72x1x1]
%onnx::Conv_609[FLOAT, 72x24x1x1]
%onnx::Conv_612[FLOAT, 72x1x5x5]
%onnx::Conv_615[FLOAT, 32x72x1x1]
%onnx::Conv_616[FLOAT, 32]
%onnx::Conv_618[FLOAT, 192x32x1x1]
%onnx::Conv_619[FLOAT, 192]
%onnx::Conv_621[FLOAT, 192x1x5x5]
%onnx::Conv_624[FLOAT, 32x192x1x1]
%onnx::Conv_627[FLOAT, 192x32x1x1]
%onnx::Conv_630[FLOAT, 192x1x3x3]
%onnx::Conv_633[FLOAT, 32x192x1x1]
%onnx::Conv_636[FLOAT, 192x32x1x1]
%onnx::Conv_639[FLOAT, 192x1x5x5]
%onnx::Conv_642[FLOAT, 32x192x1x1]
%onnx::Conv_645[FLOAT, 192x32x1x1]
%onnx::Conv_648[FLOAT, 192x1x3x3]
%onnx::Conv_651[FLOAT, 64x192x1x1]
%onnx::Conv_652[FLOAT, 64]
%onnx::Conv_654[FLOAT, 192x64x1x1]
%onnx::Conv_657[FLOAT, 192x1x3x3]
%onnx::Conv_660[FLOAT, 64x192x1x1]
%onnx::Conv_663[FLOAT, 64x64x1x1]
%onnx::Conv_666[FLOAT, 64x1x3x3]
%onnx::Conv_669[FLOAT, 64x64x1x1]
%onnx::Conv_672[FLOAT, 64x32x1x1]
%onnx::Conv_675[FLOAT, 64x1x5x5]
%onnx::Conv_678[FLOAT, 112x32x1x1]
%onnx::Conv_679[FLOAT, 112]
%onnx::Conv_681[FLOAT, 112x56x1x1]
%onnx::Conv_684[FLOAT, 112x1x5x5]
%onnx::Conv_687[FLOAT, 112x56x1x1]
%onnx::Conv_690[FLOAT, 672x112x1x1]
%onnx::Conv_691[FLOAT, 672]
%onnx::Conv_693[FLOAT, 672x1x3x3]
%onnx::Conv_696[FLOAT, 112x672x1x1]
%onnx::Conv_699[FLOAT, 184x112x1x1]
%onnx::Conv_700[FLOAT, 184]
%onnx::Conv_702[FLOAT, 552x184x1x1]
%onnx::Conv_703[FLOAT, 552]
%onnx::Conv_705[FLOAT, 552x1x3x3]
%onnx::Conv_708[FLOAT, 184x552x1x1]
%onnx::Conv_711[FLOAT, 552x184x1x1]
%onnx::Conv_714[FLOAT, 552x1x5x5]
%onnx::Conv_717[FLOAT, 184x552x1x1]
%onnx::Conv_720[FLOAT, 352x184x1x1]
%onnx::Conv_721[FLOAT, 352]
%onnx::Conv_723[FLOAT, 1504x352x1x1]
%onnx::Conv_724[FLOAT, 1504]
) {
%onnx::Conv_718 = Identity(%onnx::Conv_700)
%onnx::Conv_715 = Identity(%onnx::Conv_703)
%onnx::Conv_712 = Identity(%onnx::Conv_703)
%onnx::Conv_709 = Identity(%onnx::Conv_700)
%onnx::Conv_706 = Identity(%onnx::Conv_703)
%onnx::Conv_697 = Identity(%onnx::Conv_679)
%onnx::Conv_694 = Identity(%onnx::Conv_691)
%onnx::Conv_688 = Identity(%onnx::Conv_679)
%onnx::Conv_685 = Identity(%onnx::Conv_679)
%onnx::Conv_682 = Identity(%onnx::Conv_679)
%onnx::Conv_676 = Identity(%onnx::Conv_652)
%onnx::Conv_673 = Identity(%onnx::Conv_652)
%onnx::Conv_670 = Identity(%onnx::Conv_652)
%onnx::Conv_667 = Identity(%onnx::Conv_652)
%onnx::Conv_664 = Identity(%onnx::Conv_652)
%onnx::Conv_661 = Identity(%onnx::Conv_652)
%onnx::Conv_658 = Identity(%onnx::Conv_619)
%onnx::Conv_655 = Identity(%onnx::Conv_619)
%onnx::Conv_649 = Identity(%onnx::Conv_619)
%onnx::Conv_646 = Identity(%onnx::Conv_619)
%onnx::Conv_643 = Identity(%onnx::Conv_616)
%onnx::Conv_640 = Identity(%onnx::Conv_619)
%onnx::Conv_637 = Identity(%onnx::Conv_619)
%onnx::Conv_634 = Identity(%onnx::Conv_616)
%onnx::Conv_631 = Identity(%onnx::Conv_619)
%onnx::Conv_628 = Identity(%onnx::Conv_619)
%onnx::Conv_625 = Identity(%onnx::Conv_616)
%onnx::Conv_622 = Identity(%onnx::Conv_619)
%onnx::Conv_613 = Identity(%onnx::Conv_592)
%onnx::Conv_610 = Identity(%onnx::Conv_592)
%onnx::Conv_607 = Identity(%onnx::Conv_580)
%onnx::Conv_604 = Identity(%onnx::Conv_592)
%onnx::Conv_601 = Identity(%onnx::Conv_592)
%onnx::Conv_598 = Identity(%onnx::Conv_580)
%onnx::Conv_595 = Identity(%onnx::Conv_592)
%onnx::Conv_589 = Identity(%onnx::Conv_580)
%onnx::Conv_586 = Identity(%onnx::Conv_580)
%onnx::Conv_583 = Identity(%onnx::Conv_580)
%onnx::Conv_577 = Identity(%onnx::Conv_574)
%onnx::Conv_571 = Identity(%onnx::Conv_562)
%onnx::Conv_568 = Identity(%onnx::Conv_562)
%onnx::Conv_565 = Identity(%onnx::Conv_562)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_561, %onnx::Conv_562)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_564, %onnx::Conv_565)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_567, %onnx::Conv_568)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_570, %onnx::Conv_571)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_573, %onnx::Conv_574)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_576, %onnx::Conv_577)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_579, %onnx::Conv_580)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_582, %onnx::Conv_583)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_585, %onnx::Conv_586)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_588, %onnx::Conv_589)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_591, %onnx::Conv_592)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_594, %onnx::Conv_595)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_597, %onnx::Conv_598)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_600, %onnx::Conv_601)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_603, %onnx::Conv_604)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_606, %onnx::Conv_607)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_609, %onnx::Conv_610)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_612, %onnx::Conv_613)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_615, %onnx::Conv_616)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_618, %onnx::Conv_619)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_621, %onnx::Conv_622)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_624, %onnx::Conv_625)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_627, %onnx::Conv_628)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_630, %onnx::Conv_631)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.15/Add_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.21/relu/Relu_output_0 = Relu(%/cells.21/conv/Conv_output_0)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/relu/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%559 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %559
}
|
val_accuracy
| 0
| 66,472,960
| 1,512,980
|
{'zcp_synflow': 71.59854593809894, 'zcp_zen': 59.80867004394531, 'zcp_epe_nas': 8.57071029280668, 'zcp_fisher': 0.05603431537747383, 'zcp_flops': 66472960.0, 'zcp_grad_norm': 19.83919906616211, 'zcp_grasp': -0.0049037933349609375, 'zcp_jacov': -16.07377563476289, 'zcp_l2_norm': 536.9342041015625, 'zcp_nwot': 213.8443985392733, 'zcp_params': 1512980.0, 'zcp_plain': 0.005780612584203482, 'zcp_snip': 29.421245574951172, 'lat_1080ti_1': 0.2991702101894405, 'lat_1080ti_32': 0.35210235125572725, 'lat_1080ti_64': 0.3816471188709144, 'lat_2080ti_1': 0.35987413203952445, 'lat_2080ti_32': 0.2840470618732323, 'lat_2080ti_64': 0.3894098494012469, 'lat_essential_ph_1': 0.24528301886792453, 'lat_eyeriss': 0.434397520767242, 'lat_fpga': 0.39916286525692907, 'lat_gold_6226': 0.4175119594478486, 'lat_gold_6240': 0.2833216372751055, 'lat_pixel2': 0.21739130434782608, 'lat_pixel3': 0.43878447160638795, 'lat_raspi4': 0.3768504697470527, 'lat_samsung_a50': 0.17894736842105263, 'lat_samsung_s7': 0.11811023622047244, 'lat_silver_4114': 0.3115193908498864, 'lat_silver_4210r': 0.30586418501985707, 'lat_titan_rtx_1': 0.3295385921997984, 'lat_titan_rtx_32': 0.26759262094107017, 'lat_titan_rtx_64': 0.3458150315575669, 'lat_titanx_1': 0.17457529585144915, 'lat_titanx_32': 0.2964356938549481, 'lat_titanx_64': 0.3773717641140022, 'lat_titanxp_1': 0.31304246851186235, 'lat_titanxp_32': 0.28163801177465564, 'lat_titanxp_64': 0.38027166897715525}
| |
FBNet_4452
|
FBNet
|
4452
|
4452
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_614[FLOAT, 16x3x3x3]
%onnx::Conv_615[FLOAT, 16]
%onnx::Conv_617[FLOAT, 16x16x1x1]
%onnx::Conv_620[FLOAT, 16x1x5x5]
%onnx::Conv_623[FLOAT, 16x16x1x1]
%onnx::Conv_626[FLOAT, 96x16x1x1]
%onnx::Conv_627[FLOAT, 96]
%onnx::Conv_629[FLOAT, 96x1x5x5]
%onnx::Conv_632[FLOAT, 24x96x1x1]
%onnx::Conv_633[FLOAT, 24]
%onnx::Conv_635[FLOAT, 24x12x1x1]
%onnx::Conv_638[FLOAT, 24x1x5x5]
%onnx::Conv_641[FLOAT, 24x12x1x1]
%onnx::Conv_644[FLOAT, 72x24x1x1]
%onnx::Conv_645[FLOAT, 72]
%onnx::Conv_647[FLOAT, 72x1x3x3]
%onnx::Conv_650[FLOAT, 24x72x1x1]
%onnx::Conv_653[FLOAT, 32x24x1x1]
%onnx::Conv_654[FLOAT, 32]
%onnx::Conv_656[FLOAT, 32x32x1x1]
%onnx::Conv_659[FLOAT, 32x1x3x3]
%onnx::Conv_662[FLOAT, 32x32x1x1]
%onnx::Conv_665[FLOAT, 32x32x1x1]
%onnx::Conv_668[FLOAT, 32x1x3x3]
%onnx::Conv_671[FLOAT, 32x32x1x1]
%onnx::Conv_674[FLOAT, 32x32x1x1]
%onnx::Conv_677[FLOAT, 32x1x3x3]
%onnx::Conv_680[FLOAT, 32x32x1x1]
%onnx::Conv_683[FLOAT, 64x32x1x1]
%onnx::Conv_684[FLOAT, 64]
%onnx::Conv_686[FLOAT, 192x64x1x1]
%onnx::Conv_687[FLOAT, 192]
%onnx::Conv_689[FLOAT, 192x1x5x5]
%onnx::Conv_692[FLOAT, 64x192x1x1]
%onnx::Conv_695[FLOAT, 192x64x1x1]
%onnx::Conv_698[FLOAT, 192x1x3x3]
%onnx::Conv_701[FLOAT, 64x192x1x1]
%onnx::Conv_704[FLOAT, 192x64x1x1]
%onnx::Conv_707[FLOAT, 192x1x5x5]
%onnx::Conv_710[FLOAT, 64x192x1x1]
%onnx::Conv_713[FLOAT, 64x64x1x1]
%onnx::Conv_716[FLOAT, 64x1x3x3]
%onnx::Conv_719[FLOAT, 112x64x1x1]
%onnx::Conv_720[FLOAT, 112]
%onnx::Conv_722[FLOAT, 112x112x1x1]
%onnx::Conv_725[FLOAT, 112x1x3x3]
%onnx::Conv_728[FLOAT, 112x112x1x1]
%onnx::Conv_731[FLOAT, 672x112x1x1]
%onnx::Conv_732[FLOAT, 672]
%onnx::Conv_734[FLOAT, 672x1x3x3]
%onnx::Conv_737[FLOAT, 112x672x1x1]
%onnx::Conv_740[FLOAT, 336x112x1x1]
%onnx::Conv_741[FLOAT, 336]
%onnx::Conv_743[FLOAT, 336x1x3x3]
%onnx::Conv_746[FLOAT, 112x336x1x1]
%onnx::Conv_749[FLOAT, 336x112x1x1]
%onnx::Conv_752[FLOAT, 336x1x3x3]
%onnx::Conv_755[FLOAT, 184x336x1x1]
%onnx::Conv_756[FLOAT, 184]
%onnx::Conv_758[FLOAT, 184x92x1x1]
%onnx::Conv_761[FLOAT, 184x1x3x3]
%onnx::Conv_764[FLOAT, 184x92x1x1]
%onnx::Conv_767[FLOAT, 184x184x1x1]
%onnx::Conv_770[FLOAT, 184x1x5x5]
%onnx::Conv_773[FLOAT, 184x184x1x1]
%onnx::Conv_776[FLOAT, 1104x184x1x1]
%onnx::Conv_777[FLOAT, 1104]
%onnx::Conv_779[FLOAT, 1104x1x3x3]
%onnx::Conv_782[FLOAT, 184x1104x1x1]
%onnx::Conv_785[FLOAT, 184x92x1x1]
%onnx::Conv_788[FLOAT, 184x1x5x5]
%onnx::Conv_791[FLOAT, 352x92x1x1]
%onnx::Conv_792[FLOAT, 352]
%onnx::Conv_794[FLOAT, 1504x352x1x1]
%onnx::Conv_795[FLOAT, 1504]
) {
%onnx::Conv_789 = Identity(%onnx::Conv_756)
%onnx::Conv_786 = Identity(%onnx::Conv_756)
%onnx::Conv_783 = Identity(%onnx::Conv_756)
%onnx::Conv_780 = Identity(%onnx::Conv_777)
%onnx::Conv_774 = Identity(%onnx::Conv_756)
%onnx::Conv_771 = Identity(%onnx::Conv_756)
%onnx::Conv_768 = Identity(%onnx::Conv_756)
%onnx::Conv_765 = Identity(%onnx::Conv_756)
%onnx::Conv_762 = Identity(%onnx::Conv_756)
%onnx::Conv_759 = Identity(%onnx::Conv_756)
%onnx::Conv_753 = Identity(%onnx::Conv_741)
%onnx::Conv_750 = Identity(%onnx::Conv_741)
%onnx::Conv_747 = Identity(%onnx::Conv_720)
%onnx::Conv_744 = Identity(%onnx::Conv_741)
%onnx::Conv_738 = Identity(%onnx::Conv_720)
%onnx::Conv_735 = Identity(%onnx::Conv_732)
%onnx::Conv_729 = Identity(%onnx::Conv_720)
%onnx::Conv_726 = Identity(%onnx::Conv_720)
%onnx::Conv_723 = Identity(%onnx::Conv_720)
%onnx::Conv_717 = Identity(%onnx::Conv_684)
%onnx::Conv_714 = Identity(%onnx::Conv_684)
%onnx::Conv_711 = Identity(%onnx::Conv_684)
%onnx::Conv_708 = Identity(%onnx::Conv_687)
%onnx::Conv_705 = Identity(%onnx::Conv_687)
%onnx::Conv_702 = Identity(%onnx::Conv_684)
%onnx::Conv_699 = Identity(%onnx::Conv_687)
%onnx::Conv_696 = Identity(%onnx::Conv_687)
%onnx::Conv_693 = Identity(%onnx::Conv_684)
%onnx::Conv_690 = Identity(%onnx::Conv_687)
%onnx::Conv_681 = Identity(%onnx::Conv_654)
%onnx::Conv_678 = Identity(%onnx::Conv_654)
%onnx::Conv_675 = Identity(%onnx::Conv_654)
%onnx::Conv_672 = Identity(%onnx::Conv_654)
%onnx::Conv_669 = Identity(%onnx::Conv_654)
%onnx::Conv_666 = Identity(%onnx::Conv_654)
%onnx::Conv_663 = Identity(%onnx::Conv_654)
%onnx::Conv_660 = Identity(%onnx::Conv_654)
%onnx::Conv_657 = Identity(%onnx::Conv_654)
%onnx::Conv_651 = Identity(%onnx::Conv_633)
%onnx::Conv_648 = Identity(%onnx::Conv_645)
%onnx::Conv_642 = Identity(%onnx::Conv_633)
%onnx::Conv_639 = Identity(%onnx::Conv_633)
%onnx::Conv_636 = Identity(%onnx::Conv_633)
%onnx::Conv_630 = Identity(%onnx::Conv_627)
%onnx::Conv_624 = Identity(%onnx::Conv_615)
%onnx::Conv_621 = Identity(%onnx::Conv_615)
%onnx::Conv_618 = Identity(%onnx::Conv_615)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_614, %onnx::Conv_615)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_617, %onnx::Conv_618)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_620, %onnx::Conv_621)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_623, %onnx::Conv_624)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.4/Add_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.5/relu/Relu_output_0 = Relu(%/cells.5/conv/Conv_output_0)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/relu/Relu_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/relu/Relu_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.8/Add_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.9/relu/Relu_output_0 = Relu(%/cells.9/conv/Conv_output_0)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/relu/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/relu/Relu_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_output_0 = Reshape(%/cells.18/nl/Relu_output_0, %/cells.18/shuffle/Constant_output_0)
%/cells.18/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.18/shuffle/Reshape_output_0)
%/cells.18/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_1_output_0 = Reshape(%/cells.18/shuffle/Transpose_output_0, %/cells.18/shuffle/Constant_1_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/shuffle/Reshape_1_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%612 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %612
}
|
val_accuracy
| 0
| 61,982,336
| 1,764,156
|
{'zcp_synflow': 78.09020286557175, 'zcp_zen': 66.88166809082031, 'zcp_epe_nas': 17.09294250950314, 'zcp_fisher': 0.09789466857910156, 'zcp_flops': 61982336.0, 'zcp_grad_norm': 22.02710723876953, 'zcp_grasp': -0.06190681457519531, 'zcp_jacov': -16.06711326291932, 'zcp_l2_norm': 615.9899291992188, 'zcp_nwot': 208.63559357242968, 'zcp_params': 1764156.0, 'zcp_plain': -0.0021973969414830208, 'zcp_snip': 43.6843376159668, 'lat_1080ti_1': 0.5135680723533599, 'lat_1080ti_32': 0.5119499794511071, 'lat_1080ti_64': 0.34787677479497264, 'lat_2080ti_1': 0.5823740920755257, 'lat_2080ti_32': 0.5036076337757228, 'lat_2080ti_64': 0.35551002388394337, 'lat_essential_ph_1': 0.2641509433962264, 'lat_eyeriss': 0.33750206297104496, 'lat_fpga': 0.4011320119929942, 'lat_gold_6226': 0.29825644550612146, 'lat_gold_6240': 0.4082651663653389, 'lat_pixel2': 0.2391304347826087, 'lat_pixel3': 0.32055899550944356, 'lat_raspi4': 0.33914313108283434, 'lat_samsung_a50': 0.15789473684210525, 'lat_samsung_s7': 0.14173228346456693, 'lat_silver_4114': 0.4088774420473232, 'lat_silver_4210r': 0.4196528957247016, 'lat_titan_rtx_1': 0.5400536426266526, 'lat_titan_rtx_32': 0.48159151590554866, 'lat_titan_rtx_64': 0.3825312678002485, 'lat_titanx_1': 0.27701629160212693, 'lat_titanx_32': 0.4266960036138591, 'lat_titanx_64': 0.34160501908593777, 'lat_titanxp_1': 0.49963750368424936, 'lat_titanxp_32': 0.45530252384911624, 'lat_titanxp_64': 0.36643262048928404}
| |
FBNet_452
|
FBNet
|
452
|
452
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_732[FLOAT, 16x3x3x3]
%onnx::Conv_733[FLOAT, 16]
%onnx::Conv_735[FLOAT, 96x16x1x1]
%onnx::Conv_736[FLOAT, 96]
%onnx::Conv_738[FLOAT, 96x1x3x3]
%onnx::Conv_741[FLOAT, 16x96x1x1]
%onnx::Conv_744[FLOAT, 96x16x1x1]
%onnx::Conv_747[FLOAT, 96x1x3x3]
%onnx::Conv_750[FLOAT, 24x96x1x1]
%onnx::Conv_751[FLOAT, 24]
%onnx::Conv_753[FLOAT, 144x24x1x1]
%onnx::Conv_754[FLOAT, 144]
%onnx::Conv_756[FLOAT, 144x1x3x3]
%onnx::Conv_759[FLOAT, 24x144x1x1]
%onnx::Conv_762[FLOAT, 144x24x1x1]
%onnx::Conv_765[FLOAT, 144x1x3x3]
%onnx::Conv_768[FLOAT, 24x144x1x1]
%onnx::Conv_771[FLOAT, 72x24x1x1]
%onnx::Conv_772[FLOAT, 72]
%onnx::Conv_774[FLOAT, 72x1x3x3]
%onnx::Conv_777[FLOAT, 24x72x1x1]
%onnx::Conv_780[FLOAT, 24x12x1x1]
%onnx::Conv_783[FLOAT, 24x1x3x3]
%onnx::Conv_786[FLOAT, 32x12x1x1]
%onnx::Conv_787[FLOAT, 32]
%onnx::Conv_789[FLOAT, 32x16x1x1]
%onnx::Conv_792[FLOAT, 32x1x3x3]
%onnx::Conv_795[FLOAT, 32x16x1x1]
%onnx::Conv_798[FLOAT, 32x32x1x1]
%onnx::Conv_801[FLOAT, 32x1x3x3]
%onnx::Conv_804[FLOAT, 32x32x1x1]
%onnx::Conv_807[FLOAT, 192x32x1x1]
%onnx::Conv_808[FLOAT, 192]
%onnx::Conv_810[FLOAT, 192x1x3x3]
%onnx::Conv_813[FLOAT, 32x192x1x1]
%onnx::Conv_816[FLOAT, 96x32x1x1]
%onnx::Conv_819[FLOAT, 96x1x5x5]
%onnx::Conv_822[FLOAT, 64x96x1x1]
%onnx::Conv_823[FLOAT, 64]
%onnx::Conv_825[FLOAT, 192x64x1x1]
%onnx::Conv_828[FLOAT, 192x1x5x5]
%onnx::Conv_831[FLOAT, 64x192x1x1]
%onnx::Conv_834[FLOAT, 192x64x1x1]
%onnx::Conv_837[FLOAT, 192x1x5x5]
%onnx::Conv_840[FLOAT, 64x192x1x1]
%onnx::Conv_843[FLOAT, 64x32x1x1]
%onnx::Conv_846[FLOAT, 64x1x3x3]
%onnx::Conv_849[FLOAT, 64x32x1x1]
%onnx::Conv_852[FLOAT, 64x64x1x1]
%onnx::Conv_855[FLOAT, 64x1x3x3]
%onnx::Conv_858[FLOAT, 112x64x1x1]
%onnx::Conv_859[FLOAT, 112]
%onnx::Conv_861[FLOAT, 336x112x1x1]
%onnx::Conv_862[FLOAT, 336]
%onnx::Conv_864[FLOAT, 336x1x3x3]
%onnx::Conv_867[FLOAT, 112x336x1x1]
%onnx::Conv_870[FLOAT, 112x56x1x1]
%onnx::Conv_873[FLOAT, 112x1x5x5]
%onnx::Conv_876[FLOAT, 112x56x1x1]
%onnx::Conv_879[FLOAT, 672x112x1x1]
%onnx::Conv_880[FLOAT, 672]
%onnx::Conv_882[FLOAT, 672x1x5x5]
%onnx::Conv_885[FLOAT, 112x672x1x1]
%onnx::Conv_888[FLOAT, 112x56x1x1]
%onnx::Conv_891[FLOAT, 112x1x3x3]
%onnx::Conv_894[FLOAT, 184x56x1x1]
%onnx::Conv_895[FLOAT, 184]
%onnx::Conv_897[FLOAT, 184x92x1x1]
%onnx::Conv_900[FLOAT, 184x1x3x3]
%onnx::Conv_903[FLOAT, 184x92x1x1]
%onnx::Conv_906[FLOAT, 552x184x1x1]
%onnx::Conv_907[FLOAT, 552]
%onnx::Conv_909[FLOAT, 552x1x5x5]
%onnx::Conv_912[FLOAT, 184x552x1x1]
%onnx::Conv_915[FLOAT, 552x184x1x1]
%onnx::Conv_918[FLOAT, 552x1x5x5]
%onnx::Conv_921[FLOAT, 184x552x1x1]
%onnx::Conv_924[FLOAT, 552x184x1x1]
%onnx::Conv_927[FLOAT, 552x1x3x3]
%onnx::Conv_930[FLOAT, 352x552x1x1]
%onnx::Conv_931[FLOAT, 352]
%onnx::Conv_933[FLOAT, 1504x352x1x1]
%onnx::Conv_934[FLOAT, 1504]
) {
%onnx::Conv_928 = Identity(%onnx::Conv_907)
%onnx::Conv_925 = Identity(%onnx::Conv_907)
%onnx::Conv_922 = Identity(%onnx::Conv_895)
%onnx::Conv_919 = Identity(%onnx::Conv_907)
%onnx::Conv_916 = Identity(%onnx::Conv_907)
%onnx::Conv_913 = Identity(%onnx::Conv_895)
%onnx::Conv_910 = Identity(%onnx::Conv_907)
%onnx::Conv_904 = Identity(%onnx::Conv_895)
%onnx::Conv_901 = Identity(%onnx::Conv_895)
%onnx::Conv_898 = Identity(%onnx::Conv_895)
%onnx::Conv_892 = Identity(%onnx::Conv_859)
%onnx::Conv_889 = Identity(%onnx::Conv_859)
%onnx::Conv_886 = Identity(%onnx::Conv_859)
%onnx::Conv_883 = Identity(%onnx::Conv_880)
%onnx::Conv_877 = Identity(%onnx::Conv_859)
%onnx::Conv_874 = Identity(%onnx::Conv_859)
%onnx::Conv_871 = Identity(%onnx::Conv_859)
%onnx::Conv_868 = Identity(%onnx::Conv_859)
%onnx::Conv_865 = Identity(%onnx::Conv_862)
%onnx::Conv_856 = Identity(%onnx::Conv_823)
%onnx::Conv_853 = Identity(%onnx::Conv_823)
%onnx::Conv_850 = Identity(%onnx::Conv_823)
%onnx::Conv_847 = Identity(%onnx::Conv_823)
%onnx::Conv_844 = Identity(%onnx::Conv_823)
%onnx::Conv_841 = Identity(%onnx::Conv_823)
%onnx::Conv_838 = Identity(%onnx::Conv_808)
%onnx::Conv_835 = Identity(%onnx::Conv_808)
%onnx::Conv_832 = Identity(%onnx::Conv_823)
%onnx::Conv_829 = Identity(%onnx::Conv_808)
%onnx::Conv_826 = Identity(%onnx::Conv_808)
%onnx::Conv_820 = Identity(%onnx::Conv_736)
%onnx::Conv_817 = Identity(%onnx::Conv_736)
%onnx::Conv_814 = Identity(%onnx::Conv_787)
%onnx::Conv_811 = Identity(%onnx::Conv_808)
%onnx::Conv_805 = Identity(%onnx::Conv_787)
%onnx::Conv_802 = Identity(%onnx::Conv_787)
%onnx::Conv_799 = Identity(%onnx::Conv_787)
%onnx::Conv_796 = Identity(%onnx::Conv_787)
%onnx::Conv_793 = Identity(%onnx::Conv_787)
%onnx::Conv_790 = Identity(%onnx::Conv_787)
%onnx::Conv_784 = Identity(%onnx::Conv_751)
%onnx::Conv_781 = Identity(%onnx::Conv_751)
%onnx::Conv_778 = Identity(%onnx::Conv_751)
%onnx::Conv_775 = Identity(%onnx::Conv_772)
%onnx::Conv_769 = Identity(%onnx::Conv_751)
%onnx::Conv_766 = Identity(%onnx::Conv_754)
%onnx::Conv_763 = Identity(%onnx::Conv_754)
%onnx::Conv_760 = Identity(%onnx::Conv_751)
%onnx::Conv_757 = Identity(%onnx::Conv_754)
%onnx::Conv_748 = Identity(%onnx::Conv_736)
%onnx::Conv_745 = Identity(%onnx::Conv_736)
%onnx::Conv_742 = Identity(%onnx::Conv_733)
%onnx::Conv_739 = Identity(%onnx::Conv_736)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_732, %onnx::Conv_733)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_858, %onnx::Conv_859)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_861, %onnx::Conv_862)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_864, %onnx::Conv_865)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_867, %onnx::Conv_868)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_870, %onnx::Conv_871)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_873, %onnx::Conv_874)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_876, %onnx::Conv_877)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_879, %onnx::Conv_880)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_882, %onnx::Conv_883)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_885, %onnx::Conv_886)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_888, %onnx::Conv_889)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_891, %onnx::Conv_892)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_894, %onnx::Conv_895)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_897, %onnx::Conv_898)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_output_0 = Reshape(%/cells.18/nl/Relu_output_0, %/cells.18/shuffle/Constant_output_0)
%/cells.18/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.18/shuffle/Reshape_output_0)
%/cells.18/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_1_output_0 = Reshape(%/cells.18/shuffle/Transpose_output_0, %/cells.18/shuffle/Constant_1_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/shuffle/Reshape_1_output_0, %onnx::Conv_900, %onnx::Conv_901)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_903, %onnx::Conv_904)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_906, %onnx::Conv_907)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_909, %onnx::Conv_910)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_912, %onnx::Conv_913)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_915, %onnx::Conv_916)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_918, %onnx::Conv_919)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_921, %onnx::Conv_922)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_924, %onnx::Conv_925)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_927, %onnx::Conv_928)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_930, %onnx::Conv_931)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_933, %onnx::Conv_934)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%730 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %730
}
|
val_accuracy
| 0
| 84,194,688
| 1,888,852
|
{'zcp_synflow': 79.3659865135457, 'zcp_zen': 73.09788513183594, 'zcp_epe_nas': 7.582049734413388, 'zcp_fisher': 0.27562254667282104, 'zcp_flops': 84194688.0, 'zcp_grad_norm': 32.99165344238281, 'zcp_grasp': -0.049289703369140625, 'zcp_jacov': -16.077036062661506, 'zcp_l2_norm': 671.0658569335938, 'zcp_nwot': 219.6950144809314, 'zcp_params': 1888852.0, 'zcp_plain': 0.007460635621100664, 'zcp_snip': 58.04151153564453, 'lat_1080ti_1': 0.8231932104613646, 'lat_1080ti_32': 0.846043361553514, 'lat_1080ti_64': 0.8105911322640174, 'lat_2080ti_1': 0.8391830116530816, 'lat_2080ti_32': 0.8594677062449922, 'lat_2080ti_64': 0.853284545157523, 'lat_essential_ph_1': 0.41509433962264153, 'lat_eyeriss': 0.6336530174389817, 'lat_fpga': 0.7205394868344234, 'lat_gold_6226': 0.4007246284134942, 'lat_gold_6240': 0.6163467316050021, 'lat_pixel2': 0.41304347826086957, 'lat_pixel3': 0.5940144763721771, 'lat_raspi4': 0.6544122544662562, 'lat_samsung_a50': 0.28421052631578947, 'lat_samsung_s7': 0.29133858267716534, 'lat_silver_4114': 0.6429486495925671, 'lat_silver_4210r': 0.688155476199411, 'lat_titan_rtx_1': 0.7988584854564299, 'lat_titan_rtx_32': 0.8176139636404507, 'lat_titan_rtx_64': 0.8803629100239346, 'lat_titanx_1': 0.4244716294138091, 'lat_titanx_32': 0.8591367484422695, 'lat_titanx_64': 0.7747331497656689, 'lat_titanxp_1': 0.7428353609672214, 'lat_titanxp_32': 0.8625849880333786, 'lat_titanxp_64': 0.8267657222254122}
| |
FBNet_1677
|
FBNet
|
1677
|
1677
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_696[FLOAT, 16x3x3x3]
%onnx::Conv_697[FLOAT, 16]
%onnx::Conv_699[FLOAT, 16x16x1x1]
%onnx::Conv_702[FLOAT, 16x1x5x5]
%onnx::Conv_705[FLOAT, 16x16x1x1]
%onnx::Conv_708[FLOAT, 16x16x1x1]
%onnx::Conv_711[FLOAT, 16x1x3x3]
%onnx::Conv_714[FLOAT, 24x16x1x1]
%onnx::Conv_715[FLOAT, 24]
%onnx::Conv_717[FLOAT, 24x24x1x1]
%onnx::Conv_720[FLOAT, 24x1x5x5]
%onnx::Conv_723[FLOAT, 24x24x1x1]
%onnx::Conv_726[FLOAT, 24x12x1x1]
%onnx::Conv_729[FLOAT, 24x1x3x3]
%onnx::Conv_732[FLOAT, 24x12x1x1]
%onnx::Conv_735[FLOAT, 24x24x1x1]
%onnx::Conv_738[FLOAT, 24x1x5x5]
%onnx::Conv_741[FLOAT, 24x24x1x1]
%onnx::Conv_744[FLOAT, 32x24x1x1]
%onnx::Conv_745[FLOAT, 32]
%onnx::Conv_747[FLOAT, 96x32x1x1]
%onnx::Conv_748[FLOAT, 96]
%onnx::Conv_750[FLOAT, 96x1x5x5]
%onnx::Conv_753[FLOAT, 32x96x1x1]
%onnx::Conv_756[FLOAT, 32x16x1x1]
%onnx::Conv_759[FLOAT, 32x1x5x5]
%onnx::Conv_762[FLOAT, 32x16x1x1]
%onnx::Conv_765[FLOAT, 192x32x1x1]
%onnx::Conv_766[FLOAT, 192]
%onnx::Conv_768[FLOAT, 192x1x5x5]
%onnx::Conv_771[FLOAT, 32x192x1x1]
%onnx::Conv_774[FLOAT, 192x32x1x1]
%onnx::Conv_777[FLOAT, 192x1x3x3]
%onnx::Conv_780[FLOAT, 64x192x1x1]
%onnx::Conv_781[FLOAT, 64]
%onnx::Conv_783[FLOAT, 64x32x1x1]
%onnx::Conv_786[FLOAT, 64x1x5x5]
%onnx::Conv_789[FLOAT, 64x32x1x1]
%onnx::Conv_792[FLOAT, 64x32x1x1]
%onnx::Conv_795[FLOAT, 64x1x5x5]
%onnx::Conv_798[FLOAT, 64x32x1x1]
%onnx::Conv_801[FLOAT, 192x64x1x1]
%onnx::Conv_804[FLOAT, 192x1x3x3]
%onnx::Conv_807[FLOAT, 64x192x1x1]
%onnx::Conv_810[FLOAT, 192x64x1x1]
%onnx::Conv_813[FLOAT, 192x1x5x5]
%onnx::Conv_816[FLOAT, 112x192x1x1]
%onnx::Conv_817[FLOAT, 112]
%onnx::Conv_819[FLOAT, 672x112x1x1]
%onnx::Conv_820[FLOAT, 672]
%onnx::Conv_822[FLOAT, 672x1x3x3]
%onnx::Conv_825[FLOAT, 112x672x1x1]
%onnx::Conv_828[FLOAT, 112x112x1x1]
%onnx::Conv_831[FLOAT, 112x1x3x3]
%onnx::Conv_834[FLOAT, 112x112x1x1]
%onnx::Conv_837[FLOAT, 112x56x1x1]
%onnx::Conv_840[FLOAT, 112x1x3x3]
%onnx::Conv_843[FLOAT, 112x56x1x1]
%onnx::Conv_846[FLOAT, 112x112x1x1]
%onnx::Conv_849[FLOAT, 112x1x5x5]
%onnx::Conv_852[FLOAT, 184x112x1x1]
%onnx::Conv_853[FLOAT, 184]
%onnx::Conv_855[FLOAT, 184x184x1x1]
%onnx::Conv_858[FLOAT, 184x1x5x5]
%onnx::Conv_861[FLOAT, 184x184x1x1]
%onnx::Conv_864[FLOAT, 184x184x1x1]
%onnx::Conv_867[FLOAT, 184x1x5x5]
%onnx::Conv_870[FLOAT, 184x184x1x1]
%onnx::Conv_873[FLOAT, 552x184x1x1]
%onnx::Conv_874[FLOAT, 552]
%onnx::Conv_876[FLOAT, 552x1x5x5]
%onnx::Conv_879[FLOAT, 184x552x1x1]
%onnx::Conv_882[FLOAT, 184x184x1x1]
%onnx::Conv_885[FLOAT, 184x1x5x5]
%onnx::Conv_888[FLOAT, 352x184x1x1]
%onnx::Conv_889[FLOAT, 352]
%onnx::Conv_891[FLOAT, 1504x352x1x1]
%onnx::Conv_892[FLOAT, 1504]
) {
%onnx::Conv_886 = Identity(%onnx::Conv_853)
%onnx::Conv_883 = Identity(%onnx::Conv_853)
%onnx::Conv_880 = Identity(%onnx::Conv_853)
%onnx::Conv_877 = Identity(%onnx::Conv_874)
%onnx::Conv_871 = Identity(%onnx::Conv_853)
%onnx::Conv_868 = Identity(%onnx::Conv_853)
%onnx::Conv_865 = Identity(%onnx::Conv_853)
%onnx::Conv_862 = Identity(%onnx::Conv_853)
%onnx::Conv_859 = Identity(%onnx::Conv_853)
%onnx::Conv_856 = Identity(%onnx::Conv_853)
%onnx::Conv_850 = Identity(%onnx::Conv_817)
%onnx::Conv_847 = Identity(%onnx::Conv_817)
%onnx::Conv_844 = Identity(%onnx::Conv_817)
%onnx::Conv_841 = Identity(%onnx::Conv_817)
%onnx::Conv_838 = Identity(%onnx::Conv_817)
%onnx::Conv_835 = Identity(%onnx::Conv_817)
%onnx::Conv_832 = Identity(%onnx::Conv_817)
%onnx::Conv_829 = Identity(%onnx::Conv_817)
%onnx::Conv_826 = Identity(%onnx::Conv_817)
%onnx::Conv_823 = Identity(%onnx::Conv_820)
%onnx::Conv_814 = Identity(%onnx::Conv_766)
%onnx::Conv_811 = Identity(%onnx::Conv_766)
%onnx::Conv_808 = Identity(%onnx::Conv_781)
%onnx::Conv_805 = Identity(%onnx::Conv_766)
%onnx::Conv_802 = Identity(%onnx::Conv_766)
%onnx::Conv_799 = Identity(%onnx::Conv_781)
%onnx::Conv_796 = Identity(%onnx::Conv_781)
%onnx::Conv_793 = Identity(%onnx::Conv_781)
%onnx::Conv_790 = Identity(%onnx::Conv_781)
%onnx::Conv_787 = Identity(%onnx::Conv_781)
%onnx::Conv_784 = Identity(%onnx::Conv_781)
%onnx::Conv_778 = Identity(%onnx::Conv_766)
%onnx::Conv_775 = Identity(%onnx::Conv_766)
%onnx::Conv_772 = Identity(%onnx::Conv_745)
%onnx::Conv_769 = Identity(%onnx::Conv_766)
%onnx::Conv_763 = Identity(%onnx::Conv_745)
%onnx::Conv_760 = Identity(%onnx::Conv_745)
%onnx::Conv_757 = Identity(%onnx::Conv_745)
%onnx::Conv_754 = Identity(%onnx::Conv_745)
%onnx::Conv_751 = Identity(%onnx::Conv_748)
%onnx::Conv_742 = Identity(%onnx::Conv_715)
%onnx::Conv_739 = Identity(%onnx::Conv_715)
%onnx::Conv_736 = Identity(%onnx::Conv_715)
%onnx::Conv_733 = Identity(%onnx::Conv_715)
%onnx::Conv_730 = Identity(%onnx::Conv_715)
%onnx::Conv_727 = Identity(%onnx::Conv_715)
%onnx::Conv_724 = Identity(%onnx::Conv_715)
%onnx::Conv_721 = Identity(%onnx::Conv_715)
%onnx::Conv_718 = Identity(%onnx::Conv_715)
%onnx::Conv_712 = Identity(%onnx::Conv_697)
%onnx::Conv_709 = Identity(%onnx::Conv_697)
%onnx::Conv_706 = Identity(%onnx::Conv_697)
%onnx::Conv_703 = Identity(%onnx::Conv_697)
%onnx::Conv_700 = Identity(%onnx::Conv_697)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_696, %onnx::Conv_697)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.4/Add_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.5/relu/Relu_output_0 = Relu(%/cells.5/conv/Conv_output_0)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/relu/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/relu/Relu_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_858, %onnx::Conv_859)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_861, %onnx::Conv_862)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_864, %onnx::Conv_865)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_867, %onnx::Conv_868)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_870, %onnx::Conv_871)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_873, %onnx::Conv_874)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_876, %onnx::Conv_877)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_879, %onnx::Conv_880)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_882, %onnx::Conv_883)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_885, %onnx::Conv_886)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_888, %onnx::Conv_889)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_891, %onnx::Conv_892)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%694 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %694
}
|
val_accuracy
| 0
| 53,158,272
| 1,528,028
|
{'zcp_synflow': 82.14313480337273, 'zcp_zen': 69.88450622558594, 'zcp_epe_nas': 20.66930828800138, 'zcp_fisher': 0.15514594316482544, 'zcp_flops': 53158272.0, 'zcp_grad_norm': 24.26479721069336, 'zcp_grasp': 0.02939605712890625, 'zcp_jacov': -16.049823827493682, 'zcp_l2_norm': 608.197021484375, 'zcp_nwot': 206.12793695437492, 'zcp_params': 1528028.0, 'zcp_plain': -0.0013501675566658378, 'zcp_snip': 39.612892150878906, 'lat_1080ti_1': 0.8107474820805562, 'lat_1080ti_32': 0.5576879075356801, 'lat_1080ti_64': 0.3492530723578169, 'lat_2080ti_1': 0.7428376206105248, 'lat_2080ti_32': 0.5706024116870467, 'lat_2080ti_64': 0.34851586691813236, 'lat_essential_ph_1': 0.32075471698113206, 'lat_eyeriss': 0.2593383822639504, 'lat_fpga': 0.2730286268417459, 'lat_gold_6226': 0.1959170457813893, 'lat_gold_6240': 0.4635974193253348, 'lat_pixel2': 0.17391304347826086, 'lat_pixel3': 0.2580232074811908, 'lat_raspi4': 0.26766928346819396, 'lat_samsung_a50': 0.10526315789473684, 'lat_samsung_s7': 0.11023622047244094, 'lat_silver_4114': 0.5037490339585342, 'lat_silver_4210r': 0.5148130046538346, 'lat_titan_rtx_1': 0.7188851736838008, 'lat_titan_rtx_32': 0.587707605353737, 'lat_titan_rtx_64': 0.40197758391769245, 'lat_titanx_1': 0.38052604377679244, 'lat_titanx_32': 0.4596433183468869, 'lat_titanx_64': 0.3038344381245153, 'lat_titanxp_1': 0.6716933494253535, 'lat_titanxp_32': 0.5295861834682892, 'lat_titanxp_64': 0.34151931075784153}
| |
FBNet_1572
|
FBNet
|
1572
|
1572
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_741[FLOAT, 16x3x3x3]
%onnx::Conv_742[FLOAT, 16]
%onnx::Conv_744[FLOAT, 16x8x1x1]
%onnx::Conv_747[FLOAT, 16x1x3x3]
%onnx::Conv_750[FLOAT, 16x8x1x1]
%onnx::Conv_753[FLOAT, 96x16x1x1]
%onnx::Conv_754[FLOAT, 96]
%onnx::Conv_756[FLOAT, 96x1x3x3]
%onnx::Conv_759[FLOAT, 24x96x1x1]
%onnx::Conv_760[FLOAT, 24]
%onnx::Conv_762[FLOAT, 72x24x1x1]
%onnx::Conv_763[FLOAT, 72]
%onnx::Conv_765[FLOAT, 72x1x3x3]
%onnx::Conv_768[FLOAT, 24x72x1x1]
%onnx::Conv_771[FLOAT, 72x24x1x1]
%onnx::Conv_774[FLOAT, 72x1x5x5]
%onnx::Conv_777[FLOAT, 24x72x1x1]
%onnx::Conv_780[FLOAT, 24x24x1x1]
%onnx::Conv_783[FLOAT, 24x1x3x3]
%onnx::Conv_786[FLOAT, 24x24x1x1]
%onnx::Conv_789[FLOAT, 144x24x1x1]
%onnx::Conv_790[FLOAT, 144]
%onnx::Conv_792[FLOAT, 144x1x5x5]
%onnx::Conv_795[FLOAT, 32x144x1x1]
%onnx::Conv_796[FLOAT, 32]
%onnx::Conv_798[FLOAT, 32x16x1x1]
%onnx::Conv_801[FLOAT, 32x1x3x3]
%onnx::Conv_804[FLOAT, 32x16x1x1]
%onnx::Conv_807[FLOAT, 32x32x1x1]
%onnx::Conv_810[FLOAT, 32x1x5x5]
%onnx::Conv_813[FLOAT, 32x32x1x1]
%onnx::Conv_816[FLOAT, 32x16x1x1]
%onnx::Conv_819[FLOAT, 32x1x3x3]
%onnx::Conv_822[FLOAT, 64x16x1x1]
%onnx::Conv_823[FLOAT, 64]
%onnx::Conv_825[FLOAT, 64x32x1x1]
%onnx::Conv_828[FLOAT, 64x1x3x3]
%onnx::Conv_831[FLOAT, 64x32x1x1]
%onnx::Conv_834[FLOAT, 384x64x1x1]
%onnx::Conv_835[FLOAT, 384]
%onnx::Conv_837[FLOAT, 384x1x3x3]
%onnx::Conv_840[FLOAT, 64x384x1x1]
%onnx::Conv_843[FLOAT, 64x32x1x1]
%onnx::Conv_846[FLOAT, 64x1x3x3]
%onnx::Conv_849[FLOAT, 64x32x1x1]
%onnx::Conv_852[FLOAT, 192x64x1x1]
%onnx::Conv_853[FLOAT, 192]
%onnx::Conv_855[FLOAT, 192x1x3x3]
%onnx::Conv_858[FLOAT, 112x192x1x1]
%onnx::Conv_859[FLOAT, 112]
%onnx::Conv_861[FLOAT, 112x56x1x1]
%onnx::Conv_864[FLOAT, 112x1x5x5]
%onnx::Conv_867[FLOAT, 112x56x1x1]
%onnx::Conv_870[FLOAT, 112x112x1x1]
%onnx::Conv_873[FLOAT, 112x1x5x5]
%onnx::Conv_876[FLOAT, 112x112x1x1]
%onnx::Conv_879[FLOAT, 112x56x1x1]
%onnx::Conv_882[FLOAT, 112x1x5x5]
%onnx::Conv_885[FLOAT, 112x56x1x1]
%onnx::Conv_888[FLOAT, 112x112x1x1]
%onnx::Conv_891[FLOAT, 112x1x5x5]
%onnx::Conv_894[FLOAT, 184x112x1x1]
%onnx::Conv_895[FLOAT, 184]
%onnx::Conv_897[FLOAT, 184x184x1x1]
%onnx::Conv_900[FLOAT, 184x1x5x5]
%onnx::Conv_903[FLOAT, 184x184x1x1]
%onnx::Conv_906[FLOAT, 184x92x1x1]
%onnx::Conv_909[FLOAT, 184x1x5x5]
%onnx::Conv_912[FLOAT, 184x92x1x1]
%onnx::Conv_915[FLOAT, 1104x184x1x1]
%onnx::Conv_916[FLOAT, 1104]
%onnx::Conv_918[FLOAT, 1104x1x3x3]
%onnx::Conv_921[FLOAT, 184x1104x1x1]
%onnx::Conv_924[FLOAT, 184x184x1x1]
%onnx::Conv_927[FLOAT, 184x1x3x3]
%onnx::Conv_930[FLOAT, 352x184x1x1]
%onnx::Conv_931[FLOAT, 352]
%onnx::Conv_933[FLOAT, 1504x352x1x1]
%onnx::Conv_934[FLOAT, 1504]
) {
%onnx::Conv_928 = Identity(%onnx::Conv_895)
%onnx::Conv_925 = Identity(%onnx::Conv_895)
%onnx::Conv_922 = Identity(%onnx::Conv_895)
%onnx::Conv_919 = Identity(%onnx::Conv_916)
%onnx::Conv_913 = Identity(%onnx::Conv_895)
%onnx::Conv_910 = Identity(%onnx::Conv_895)
%onnx::Conv_907 = Identity(%onnx::Conv_895)
%onnx::Conv_904 = Identity(%onnx::Conv_895)
%onnx::Conv_901 = Identity(%onnx::Conv_895)
%onnx::Conv_898 = Identity(%onnx::Conv_895)
%onnx::Conv_892 = Identity(%onnx::Conv_859)
%onnx::Conv_889 = Identity(%onnx::Conv_859)
%onnx::Conv_886 = Identity(%onnx::Conv_859)
%onnx::Conv_883 = Identity(%onnx::Conv_859)
%onnx::Conv_880 = Identity(%onnx::Conv_859)
%onnx::Conv_877 = Identity(%onnx::Conv_859)
%onnx::Conv_874 = Identity(%onnx::Conv_859)
%onnx::Conv_871 = Identity(%onnx::Conv_859)
%onnx::Conv_868 = Identity(%onnx::Conv_859)
%onnx::Conv_865 = Identity(%onnx::Conv_859)
%onnx::Conv_862 = Identity(%onnx::Conv_859)
%onnx::Conv_856 = Identity(%onnx::Conv_853)
%onnx::Conv_850 = Identity(%onnx::Conv_823)
%onnx::Conv_847 = Identity(%onnx::Conv_823)
%onnx::Conv_844 = Identity(%onnx::Conv_823)
%onnx::Conv_841 = Identity(%onnx::Conv_823)
%onnx::Conv_838 = Identity(%onnx::Conv_835)
%onnx::Conv_832 = Identity(%onnx::Conv_823)
%onnx::Conv_829 = Identity(%onnx::Conv_823)
%onnx::Conv_826 = Identity(%onnx::Conv_823)
%onnx::Conv_820 = Identity(%onnx::Conv_796)
%onnx::Conv_817 = Identity(%onnx::Conv_796)
%onnx::Conv_814 = Identity(%onnx::Conv_796)
%onnx::Conv_811 = Identity(%onnx::Conv_796)
%onnx::Conv_808 = Identity(%onnx::Conv_796)
%onnx::Conv_805 = Identity(%onnx::Conv_796)
%onnx::Conv_802 = Identity(%onnx::Conv_796)
%onnx::Conv_799 = Identity(%onnx::Conv_796)
%onnx::Conv_793 = Identity(%onnx::Conv_790)
%onnx::Conv_787 = Identity(%onnx::Conv_760)
%onnx::Conv_784 = Identity(%onnx::Conv_760)
%onnx::Conv_781 = Identity(%onnx::Conv_760)
%onnx::Conv_778 = Identity(%onnx::Conv_760)
%onnx::Conv_775 = Identity(%onnx::Conv_763)
%onnx::Conv_772 = Identity(%onnx::Conv_763)
%onnx::Conv_769 = Identity(%onnx::Conv_760)
%onnx::Conv_766 = Identity(%onnx::Conv_763)
%onnx::Conv_757 = Identity(%onnx::Conv_754)
%onnx::Conv_751 = Identity(%onnx::Conv_742)
%onnx::Conv_748 = Identity(%onnx::Conv_742)
%onnx::Conv_745 = Identity(%onnx::Conv_742)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_741, %onnx::Conv_742)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_858, %onnx::Conv_859)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_861, %onnx::Conv_862)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_864, %onnx::Conv_865)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_867, %onnx::Conv_868)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_870, %onnx::Conv_871)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_873, %onnx::Conv_874)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_876, %onnx::Conv_877)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_879, %onnx::Conv_880)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_882, %onnx::Conv_883)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_885, %onnx::Conv_886)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_888, %onnx::Conv_889)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_891, %onnx::Conv_892)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_894, %onnx::Conv_895)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_897, %onnx::Conv_898)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_900, %onnx::Conv_901)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_903, %onnx::Conv_904)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_906, %onnx::Conv_907)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_909, %onnx::Conv_910)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_912, %onnx::Conv_913)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_915, %onnx::Conv_916)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_918, %onnx::Conv_919)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_921, %onnx::Conv_922)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_924, %onnx::Conv_925)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_927, %onnx::Conv_928)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_930, %onnx::Conv_931)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_933, %onnx::Conv_934)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%739 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %739
}
|
val_accuracy
| 0
| 56,373,376
| 1,553,428
|
{'zcp_synflow': 75.0401621825422, 'zcp_zen': 66.38383483886719, 'zcp_epe_nas': 10.330459141673868, 'zcp_fisher': 0.1004803329706192, 'zcp_flops': 56373376.0, 'zcp_grad_norm': 25.71776580810547, 'zcp_grasp': -0.03828620910644531, 'zcp_jacov': -16.049961398328733, 'zcp_l2_norm': 574.4906005859375, 'zcp_nwot': 212.2335093308847, 'zcp_params': 1553428.0, 'zcp_plain': 0.009278560988605022, 'zcp_snip': 43.063663482666016, 'lat_1080ti_1': 0.7001199122133086, 'lat_1080ti_32': 0.7237544574424905, 'lat_1080ti_64': 0.5262826696126183, 'lat_2080ti_1': 0.7712801906397132, 'lat_2080ti_32': 0.8672884249610164, 'lat_2080ti_64': 0.5756735586441711, 'lat_essential_ph_1': 0.3584905660377358, 'lat_eyeriss': 0.32450992976729703, 'lat_fpga': 0.29879575297598426, 'lat_gold_6226': 0.19169144174873193, 'lat_gold_6240': 0.4482234404447136, 'lat_pixel2': 0.2391304347826087, 'lat_pixel3': 0.31266434679422905, 'lat_raspi4': 0.34341846661650094, 'lat_samsung_a50': 0.14736842105263157, 'lat_samsung_s7': 0.13385826771653545, 'lat_silver_4114': 0.43190987614573617, 'lat_silver_4210r': 0.4718273072858383, 'lat_titan_rtx_1': 0.7465543906240415, 'lat_titan_rtx_32': 0.6767669607171207, 'lat_titan_rtx_64': 0.6397693332604211, 'lat_titanx_1': 0.3899240482945128, 'lat_titanx_32': 0.6960920796117207, 'lat_titanx_64': 0.522447057355192, 'lat_titanxp_1': 0.685648828564188, 'lat_titanxp_32': 0.6736802980129191, 'lat_titanxp_64': 0.5586102216824145}
| |
FBNet_915
|
FBNet
|
915
|
915
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_685[FLOAT, 16x3x3x3]
%onnx::Conv_686[FLOAT, 16]
%onnx::Conv_688[FLOAT, 48x16x1x1]
%onnx::Conv_689[FLOAT, 48]
%onnx::Conv_691[FLOAT, 48x1x5x5]
%onnx::Conv_694[FLOAT, 16x48x1x1]
%onnx::Conv_697[FLOAT, 96x16x1x1]
%onnx::Conv_698[FLOAT, 96]
%onnx::Conv_700[FLOAT, 96x1x5x5]
%onnx::Conv_703[FLOAT, 24x96x1x1]
%onnx::Conv_704[FLOAT, 24]
%onnx::Conv_706[FLOAT, 24x12x1x1]
%onnx::Conv_709[FLOAT, 24x1x5x5]
%onnx::Conv_712[FLOAT, 24x12x1x1]
%onnx::Conv_715[FLOAT, 24x12x1x1]
%onnx::Conv_718[FLOAT, 24x1x5x5]
%onnx::Conv_721[FLOAT, 24x12x1x1]
%onnx::Conv_724[FLOAT, 24x24x1x1]
%onnx::Conv_727[FLOAT, 24x1x3x3]
%onnx::Conv_730[FLOAT, 24x24x1x1]
%onnx::Conv_733[FLOAT, 24x24x1x1]
%onnx::Conv_736[FLOAT, 24x1x5x5]
%onnx::Conv_739[FLOAT, 32x24x1x1]
%onnx::Conv_740[FLOAT, 32]
%onnx::Conv_742[FLOAT, 96x32x1x1]
%onnx::Conv_745[FLOAT, 96x1x3x3]
%onnx::Conv_748[FLOAT, 32x96x1x1]
%onnx::Conv_751[FLOAT, 32x32x1x1]
%onnx::Conv_754[FLOAT, 32x1x5x5]
%onnx::Conv_757[FLOAT, 32x32x1x1]
%onnx::Conv_760[FLOAT, 96x32x1x1]
%onnx::Conv_763[FLOAT, 96x1x3x3]
%onnx::Conv_766[FLOAT, 32x96x1x1]
%onnx::Conv_769[FLOAT, 96x32x1x1]
%onnx::Conv_772[FLOAT, 96x1x5x5]
%onnx::Conv_775[FLOAT, 64x96x1x1]
%onnx::Conv_776[FLOAT, 64]
%onnx::Conv_778[FLOAT, 64x32x1x1]
%onnx::Conv_781[FLOAT, 64x1x5x5]
%onnx::Conv_784[FLOAT, 64x32x1x1]
%onnx::Conv_787[FLOAT, 64x64x1x1]
%onnx::Conv_790[FLOAT, 64x1x5x5]
%onnx::Conv_793[FLOAT, 64x64x1x1]
%onnx::Conv_796[FLOAT, 64x64x1x1]
%onnx::Conv_799[FLOAT, 64x1x3x3]
%onnx::Conv_802[FLOAT, 112x64x1x1]
%onnx::Conv_803[FLOAT, 112]
%onnx::Conv_805[FLOAT, 112x112x1x1]
%onnx::Conv_808[FLOAT, 112x1x3x3]
%onnx::Conv_811[FLOAT, 112x112x1x1]
%onnx::Conv_814[FLOAT, 336x112x1x1]
%onnx::Conv_815[FLOAT, 336]
%onnx::Conv_817[FLOAT, 336x1x3x3]
%onnx::Conv_820[FLOAT, 112x336x1x1]
%onnx::Conv_823[FLOAT, 112x112x1x1]
%onnx::Conv_826[FLOAT, 112x1x5x5]
%onnx::Conv_829[FLOAT, 112x112x1x1]
%onnx::Conv_832[FLOAT, 112x112x1x1]
%onnx::Conv_835[FLOAT, 112x1x3x3]
%onnx::Conv_838[FLOAT, 184x112x1x1]
%onnx::Conv_839[FLOAT, 184]
%onnx::Conv_841[FLOAT, 184x184x1x1]
%onnx::Conv_844[FLOAT, 184x1x3x3]
%onnx::Conv_847[FLOAT, 184x184x1x1]
%onnx::Conv_850[FLOAT, 552x184x1x1]
%onnx::Conv_851[FLOAT, 552]
%onnx::Conv_853[FLOAT, 552x1x5x5]
%onnx::Conv_856[FLOAT, 184x552x1x1]
%onnx::Conv_859[FLOAT, 184x92x1x1]
%onnx::Conv_862[FLOAT, 184x1x3x3]
%onnx::Conv_865[FLOAT, 184x92x1x1]
%onnx::Conv_868[FLOAT, 184x92x1x1]
%onnx::Conv_871[FLOAT, 184x1x5x5]
%onnx::Conv_874[FLOAT, 352x92x1x1]
%onnx::Conv_875[FLOAT, 352]
%onnx::Conv_877[FLOAT, 1504x352x1x1]
%onnx::Conv_878[FLOAT, 1504]
) {
%onnx::Conv_872 = Identity(%onnx::Conv_839)
%onnx::Conv_869 = Identity(%onnx::Conv_839)
%onnx::Conv_866 = Identity(%onnx::Conv_839)
%onnx::Conv_863 = Identity(%onnx::Conv_839)
%onnx::Conv_860 = Identity(%onnx::Conv_839)
%onnx::Conv_857 = Identity(%onnx::Conv_839)
%onnx::Conv_854 = Identity(%onnx::Conv_851)
%onnx::Conv_848 = Identity(%onnx::Conv_839)
%onnx::Conv_845 = Identity(%onnx::Conv_839)
%onnx::Conv_842 = Identity(%onnx::Conv_839)
%onnx::Conv_836 = Identity(%onnx::Conv_803)
%onnx::Conv_833 = Identity(%onnx::Conv_803)
%onnx::Conv_830 = Identity(%onnx::Conv_803)
%onnx::Conv_827 = Identity(%onnx::Conv_803)
%onnx::Conv_824 = Identity(%onnx::Conv_803)
%onnx::Conv_821 = Identity(%onnx::Conv_803)
%onnx::Conv_818 = Identity(%onnx::Conv_815)
%onnx::Conv_812 = Identity(%onnx::Conv_803)
%onnx::Conv_809 = Identity(%onnx::Conv_803)
%onnx::Conv_806 = Identity(%onnx::Conv_803)
%onnx::Conv_800 = Identity(%onnx::Conv_776)
%onnx::Conv_797 = Identity(%onnx::Conv_776)
%onnx::Conv_794 = Identity(%onnx::Conv_776)
%onnx::Conv_791 = Identity(%onnx::Conv_776)
%onnx::Conv_788 = Identity(%onnx::Conv_776)
%onnx::Conv_785 = Identity(%onnx::Conv_776)
%onnx::Conv_782 = Identity(%onnx::Conv_776)
%onnx::Conv_779 = Identity(%onnx::Conv_776)
%onnx::Conv_773 = Identity(%onnx::Conv_698)
%onnx::Conv_770 = Identity(%onnx::Conv_698)
%onnx::Conv_767 = Identity(%onnx::Conv_740)
%onnx::Conv_764 = Identity(%onnx::Conv_698)
%onnx::Conv_761 = Identity(%onnx::Conv_698)
%onnx::Conv_758 = Identity(%onnx::Conv_740)
%onnx::Conv_755 = Identity(%onnx::Conv_740)
%onnx::Conv_752 = Identity(%onnx::Conv_740)
%onnx::Conv_749 = Identity(%onnx::Conv_740)
%onnx::Conv_746 = Identity(%onnx::Conv_698)
%onnx::Conv_743 = Identity(%onnx::Conv_698)
%onnx::Conv_737 = Identity(%onnx::Conv_704)
%onnx::Conv_734 = Identity(%onnx::Conv_704)
%onnx::Conv_731 = Identity(%onnx::Conv_704)
%onnx::Conv_728 = Identity(%onnx::Conv_704)
%onnx::Conv_725 = Identity(%onnx::Conv_704)
%onnx::Conv_722 = Identity(%onnx::Conv_704)
%onnx::Conv_719 = Identity(%onnx::Conv_704)
%onnx::Conv_716 = Identity(%onnx::Conv_704)
%onnx::Conv_713 = Identity(%onnx::Conv_704)
%onnx::Conv_710 = Identity(%onnx::Conv_704)
%onnx::Conv_707 = Identity(%onnx::Conv_704)
%onnx::Conv_701 = Identity(%onnx::Conv_698)
%onnx::Conv_695 = Identity(%onnx::Conv_686)
%onnx::Conv_692 = Identity(%onnx::Conv_689)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_685, %onnx::Conv_686)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_823, %onnx::Conv_824)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_826, %onnx::Conv_827)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_829, %onnx::Conv_830)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_832, %onnx::Conv_833)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_835, %onnx::Conv_836)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_838, %onnx::Conv_839)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_841, %onnx::Conv_842)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_844, %onnx::Conv_845)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_847, %onnx::Conv_848)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_850, %onnx::Conv_851)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_853, %onnx::Conv_854)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_856, %onnx::Conv_857)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_859, %onnx::Conv_860)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_862, %onnx::Conv_863)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_865, %onnx::Conv_866)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_868, %onnx::Conv_869)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_871, %onnx::Conv_872)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_874, %onnx::Conv_875)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_877, %onnx::Conv_878)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%683 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %683
}
|
val_accuracy
| 0
| 48,462,720
| 1,310,036
|
{'zcp_synflow': 79.9398589463383, 'zcp_zen': 67.3848648071289, 'zcp_epe_nas': 8.575353070086553, 'zcp_fisher': 0.1938859522342682, 'zcp_flops': 48462720.0, 'zcp_grad_norm': 29.532739639282227, 'zcp_grasp': -0.19106674194335938, 'zcp_jacov': -16.06110850493377, 'zcp_l2_norm': 559.4659423828125, 'zcp_nwot': 208.02673405951327, 'zcp_params': 1310036.0, 'zcp_plain': 0.0053227306343615055, 'zcp_snip': 50.75046157836914, 'lat_1080ti_1': 0.6676213662648045, 'lat_1080ti_32': 0.5604326390499813, 'lat_1080ti_64': 0.39847253932218685, 'lat_2080ti_1': 0.6949359348435954, 'lat_2080ti_32': 0.5728654374610832, 'lat_2080ti_64': 0.40060675264022216, 'lat_essential_ph_1': 0.2641509433962264, 'lat_eyeriss': 0.23999229824143192, 'lat_fpga': 0.2032080269941321, 'lat_gold_6226': 0.11618624319525521, 'lat_gold_6240': 0.41409433524161054, 'lat_pixel2': 0.21739130434782608, 'lat_pixel3': 0.2631249310777275, 'lat_raspi4': 0.2616350604633028, 'lat_samsung_a50': 0.09473684210526316, 'lat_samsung_s7': 0.08661417322834646, 'lat_silver_4114': 0.5086060581659771, 'lat_silver_4210r': 0.5071091457997806, 'lat_titan_rtx_1': 0.6726828908640959, 'lat_titan_rtx_32': 0.57978173509057, 'lat_titan_rtx_64': 0.4649562519644561, 'lat_titanx_1': 0.3599558980092606, 'lat_titanx_32': 0.5177920748789753, 'lat_titanx_64': 0.3863738244703887, 'lat_titanxp_1': 0.6321426774395323, 'lat_titanxp_32': 0.5740848368555664, 'lat_titanxp_64': 0.42819240436386863}
| |
FBNet_4586
|
FBNet
|
4586
|
4586
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_749[FLOAT, 16x3x3x3]
%onnx::Conv_750[FLOAT, 16]
%onnx::Conv_752[FLOAT, 48x16x1x1]
%onnx::Conv_753[FLOAT, 48]
%onnx::Conv_755[FLOAT, 48x1x5x5]
%onnx::Conv_758[FLOAT, 16x48x1x1]
%onnx::Conv_761[FLOAT, 48x16x1x1]
%onnx::Conv_764[FLOAT, 48x1x3x3]
%onnx::Conv_767[FLOAT, 24x48x1x1]
%onnx::Conv_768[FLOAT, 24]
%onnx::Conv_770[FLOAT, 144x24x1x1]
%onnx::Conv_771[FLOAT, 144]
%onnx::Conv_773[FLOAT, 144x1x5x5]
%onnx::Conv_776[FLOAT, 24x144x1x1]
%onnx::Conv_779[FLOAT, 144x24x1x1]
%onnx::Conv_782[FLOAT, 144x1x3x3]
%onnx::Conv_785[FLOAT, 24x144x1x1]
%onnx::Conv_788[FLOAT, 144x24x1x1]
%onnx::Conv_791[FLOAT, 144x1x3x3]
%onnx::Conv_794[FLOAT, 24x144x1x1]
%onnx::Conv_797[FLOAT, 144x24x1x1]
%onnx::Conv_800[FLOAT, 144x1x5x5]
%onnx::Conv_803[FLOAT, 32x144x1x1]
%onnx::Conv_804[FLOAT, 32]
%onnx::Conv_806[FLOAT, 32x32x1x1]
%onnx::Conv_809[FLOAT, 32x1x5x5]
%onnx::Conv_812[FLOAT, 32x32x1x1]
%onnx::Conv_815[FLOAT, 32x16x1x1]
%onnx::Conv_818[FLOAT, 32x1x3x3]
%onnx::Conv_821[FLOAT, 32x16x1x1]
%onnx::Conv_824[FLOAT, 32x16x1x1]
%onnx::Conv_827[FLOAT, 32x1x5x5]
%onnx::Conv_830[FLOAT, 32x16x1x1]
%onnx::Conv_833[FLOAT, 32x16x1x1]
%onnx::Conv_836[FLOAT, 32x1x3x3]
%onnx::Conv_839[FLOAT, 64x16x1x1]
%onnx::Conv_840[FLOAT, 64]
%onnx::Conv_842[FLOAT, 64x64x1x1]
%onnx::Conv_845[FLOAT, 64x1x5x5]
%onnx::Conv_848[FLOAT, 64x64x1x1]
%onnx::Conv_851[FLOAT, 384x64x1x1]
%onnx::Conv_852[FLOAT, 384]
%onnx::Conv_854[FLOAT, 384x1x3x3]
%onnx::Conv_857[FLOAT, 64x384x1x1]
%onnx::Conv_860[FLOAT, 64x32x1x1]
%onnx::Conv_863[FLOAT, 64x1x3x3]
%onnx::Conv_866[FLOAT, 64x32x1x1]
%onnx::Conv_869[FLOAT, 384x64x1x1]
%onnx::Conv_872[FLOAT, 384x1x3x3]
%onnx::Conv_875[FLOAT, 112x384x1x1]
%onnx::Conv_876[FLOAT, 112]
%onnx::Conv_878[FLOAT, 112x56x1x1]
%onnx::Conv_881[FLOAT, 112x1x3x3]
%onnx::Conv_884[FLOAT, 112x56x1x1]
%onnx::Conv_887[FLOAT, 336x112x1x1]
%onnx::Conv_888[FLOAT, 336]
%onnx::Conv_890[FLOAT, 336x1x5x5]
%onnx::Conv_893[FLOAT, 112x336x1x1]
%onnx::Conv_896[FLOAT, 672x112x1x1]
%onnx::Conv_897[FLOAT, 672]
%onnx::Conv_899[FLOAT, 672x1x3x3]
%onnx::Conv_902[FLOAT, 112x672x1x1]
%onnx::Conv_905[FLOAT, 112x56x1x1]
%onnx::Conv_908[FLOAT, 112x1x3x3]
%onnx::Conv_911[FLOAT, 184x56x1x1]
%onnx::Conv_912[FLOAT, 184]
%onnx::Conv_914[FLOAT, 184x184x1x1]
%onnx::Conv_917[FLOAT, 184x1x3x3]
%onnx::Conv_920[FLOAT, 184x184x1x1]
%onnx::Conv_923[FLOAT, 184x92x1x1]
%onnx::Conv_926[FLOAT, 184x1x5x5]
%onnx::Conv_929[FLOAT, 184x92x1x1]
%onnx::Conv_932[FLOAT, 552x184x1x1]
%onnx::Conv_933[FLOAT, 552]
%onnx::Conv_935[FLOAT, 552x1x5x5]
%onnx::Conv_938[FLOAT, 184x552x1x1]
%onnx::Conv_941[FLOAT, 1104x184x1x1]
%onnx::Conv_942[FLOAT, 1104]
%onnx::Conv_944[FLOAT, 1104x1x5x5]
%onnx::Conv_947[FLOAT, 352x1104x1x1]
%onnx::Conv_948[FLOAT, 352]
%onnx::Conv_950[FLOAT, 1504x352x1x1]
%onnx::Conv_951[FLOAT, 1504]
) {
%onnx::Conv_945 = Identity(%onnx::Conv_942)
%onnx::Conv_939 = Identity(%onnx::Conv_912)
%onnx::Conv_936 = Identity(%onnx::Conv_933)
%onnx::Conv_930 = Identity(%onnx::Conv_912)
%onnx::Conv_927 = Identity(%onnx::Conv_912)
%onnx::Conv_924 = Identity(%onnx::Conv_912)
%onnx::Conv_921 = Identity(%onnx::Conv_912)
%onnx::Conv_918 = Identity(%onnx::Conv_912)
%onnx::Conv_915 = Identity(%onnx::Conv_912)
%onnx::Conv_909 = Identity(%onnx::Conv_876)
%onnx::Conv_906 = Identity(%onnx::Conv_876)
%onnx::Conv_903 = Identity(%onnx::Conv_876)
%onnx::Conv_900 = Identity(%onnx::Conv_897)
%onnx::Conv_894 = Identity(%onnx::Conv_876)
%onnx::Conv_891 = Identity(%onnx::Conv_888)
%onnx::Conv_885 = Identity(%onnx::Conv_876)
%onnx::Conv_882 = Identity(%onnx::Conv_876)
%onnx::Conv_879 = Identity(%onnx::Conv_876)
%onnx::Conv_873 = Identity(%onnx::Conv_852)
%onnx::Conv_870 = Identity(%onnx::Conv_852)
%onnx::Conv_867 = Identity(%onnx::Conv_840)
%onnx::Conv_864 = Identity(%onnx::Conv_840)
%onnx::Conv_861 = Identity(%onnx::Conv_840)
%onnx::Conv_858 = Identity(%onnx::Conv_840)
%onnx::Conv_855 = Identity(%onnx::Conv_852)
%onnx::Conv_849 = Identity(%onnx::Conv_840)
%onnx::Conv_846 = Identity(%onnx::Conv_840)
%onnx::Conv_843 = Identity(%onnx::Conv_840)
%onnx::Conv_837 = Identity(%onnx::Conv_804)
%onnx::Conv_834 = Identity(%onnx::Conv_804)
%onnx::Conv_831 = Identity(%onnx::Conv_804)
%onnx::Conv_828 = Identity(%onnx::Conv_804)
%onnx::Conv_825 = Identity(%onnx::Conv_804)
%onnx::Conv_822 = Identity(%onnx::Conv_804)
%onnx::Conv_819 = Identity(%onnx::Conv_804)
%onnx::Conv_816 = Identity(%onnx::Conv_804)
%onnx::Conv_813 = Identity(%onnx::Conv_804)
%onnx::Conv_810 = Identity(%onnx::Conv_804)
%onnx::Conv_807 = Identity(%onnx::Conv_804)
%onnx::Conv_801 = Identity(%onnx::Conv_771)
%onnx::Conv_798 = Identity(%onnx::Conv_771)
%onnx::Conv_795 = Identity(%onnx::Conv_768)
%onnx::Conv_792 = Identity(%onnx::Conv_771)
%onnx::Conv_789 = Identity(%onnx::Conv_771)
%onnx::Conv_786 = Identity(%onnx::Conv_768)
%onnx::Conv_783 = Identity(%onnx::Conv_771)
%onnx::Conv_780 = Identity(%onnx::Conv_771)
%onnx::Conv_777 = Identity(%onnx::Conv_768)
%onnx::Conv_774 = Identity(%onnx::Conv_771)
%onnx::Conv_765 = Identity(%onnx::Conv_753)
%onnx::Conv_762 = Identity(%onnx::Conv_753)
%onnx::Conv_759 = Identity(%onnx::Conv_750)
%onnx::Conv_756 = Identity(%onnx::Conv_753)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_749, %onnx::Conv_750)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_803, %onnx::Conv_804)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_806, %onnx::Conv_807)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_809, %onnx::Conv_810)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_812, %onnx::Conv_813)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_815, %onnx::Conv_816)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_818, %onnx::Conv_819)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_821, %onnx::Conv_822)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_824, %onnx::Conv_825)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_827, %onnx::Conv_828)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_830, %onnx::Conv_831)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_833, %onnx::Conv_834)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_836, %onnx::Conv_837)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_839, %onnx::Conv_840)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_842, %onnx::Conv_843)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_845, %onnx::Conv_846)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_848, %onnx::Conv_849)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_851, %onnx::Conv_852)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_854, %onnx::Conv_855)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_857, %onnx::Conv_858)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_860, %onnx::Conv_861)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_863, %onnx::Conv_864)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_866, %onnx::Conv_867)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_869, %onnx::Conv_870)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_872, %onnx::Conv_873)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_875, %onnx::Conv_876)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_878, %onnx::Conv_879)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_881, %onnx::Conv_882)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_884, %onnx::Conv_885)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_887, %onnx::Conv_888)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_890, %onnx::Conv_891)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_893, %onnx::Conv_894)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_896, %onnx::Conv_897)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_899, %onnx::Conv_900)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_902, %onnx::Conv_903)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_905, %onnx::Conv_906)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_908, %onnx::Conv_909)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_911, %onnx::Conv_912)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_914, %onnx::Conv_915)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_917, %onnx::Conv_918)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_920, %onnx::Conv_921)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_923, %onnx::Conv_924)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_926, %onnx::Conv_927)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_929, %onnx::Conv_930)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_932, %onnx::Conv_933)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_935, %onnx::Conv_936)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_938, %onnx::Conv_939)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_941, %onnx::Conv_942)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_944, %onnx::Conv_945)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_947, %onnx::Conv_948)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_950, %onnx::Conv_951)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%747 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %747
}
|
val_accuracy
| 0
| 94,651,008
| 2,112,332
|
{'zcp_synflow': 79.53707617227376, 'zcp_zen': 72.84576416015625, 'zcp_epe_nas': 22.394333312612048, 'zcp_fisher': 0.17721125483512878, 'zcp_flops': 94651008.0, 'zcp_grad_norm': 32.47683334350586, 'zcp_grasp': -0.02448272705078125, 'zcp_jacov': -16.071550348752197, 'zcp_l2_norm': 672.772216796875, 'zcp_nwot': 220.15405461275074, 'zcp_params': 2112332.0, 'zcp_plain': 0.00035573210334405303, 'zcp_snip': 62.60493469238281, 'lat_1080ti_1': 0.7943158781193942, 'lat_1080ti_32': 0.9037600754018819, 'lat_1080ti_64': 0.939565604167619, 'lat_2080ti_1': 0.8871160882635244, 'lat_2080ti_32': 0.9844074582810957, 'lat_2080ti_64': 0.9814404697371159, 'lat_essential_ph_1': 0.4716981132075472, 'lat_eyeriss': 0.7411750683072634, 'lat_fpga': 0.8012448173839046, 'lat_gold_6226': 0.46395923157517904, 'lat_gold_6240': 0.61209898784753, 'lat_pixel2': 0.45652173913043476, 'lat_pixel3': 0.7576114454446992, 'lat_raspi4': 0.8261370920353418, 'lat_samsung_a50': 0.30526315789473685, 'lat_samsung_s7': 0.25196850393700787, 'lat_silver_4114': 0.6383514569372261, 'lat_silver_4210r': 0.7009677431512003, 'lat_titan_rtx_1': 0.8332464357781053, 'lat_titan_rtx_32': 0.9041226590878849, 'lat_titan_rtx_64': 0.986222164174496, 'lat_titanx_1': 0.4490545291401406, 'lat_titanx_32': 0.9652460603489424, 'lat_titanx_64': 0.9639418774609616, 'lat_titanxp_1': 0.7814694906660868, 'lat_titanxp_32': 0.9591759170303612, 'lat_titanxp_64': 0.9400815533885111}
| |
FBNet_1250
|
FBNet
|
1250
|
1250
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_584[FLOAT, 16x3x3x3]
%onnx::Conv_585[FLOAT, 16]
%onnx::Conv_587[FLOAT, 96x16x1x1]
%onnx::Conv_588[FLOAT, 96]
%onnx::Conv_590[FLOAT, 96x1x5x5]
%onnx::Conv_593[FLOAT, 16x96x1x1]
%onnx::Conv_596[FLOAT, 16x16x1x1]
%onnx::Conv_599[FLOAT, 16x1x3x3]
%onnx::Conv_602[FLOAT, 24x16x1x1]
%onnx::Conv_603[FLOAT, 24]
%onnx::Conv_605[FLOAT, 24x24x1x1]
%onnx::Conv_608[FLOAT, 24x1x5x5]
%onnx::Conv_611[FLOAT, 24x24x1x1]
%onnx::Conv_614[FLOAT, 72x24x1x1]
%onnx::Conv_615[FLOAT, 72]
%onnx::Conv_617[FLOAT, 72x1x3x3]
%onnx::Conv_620[FLOAT, 24x72x1x1]
%onnx::Conv_623[FLOAT, 72x24x1x1]
%onnx::Conv_626[FLOAT, 72x1x3x3]
%onnx::Conv_629[FLOAT, 24x72x1x1]
%onnx::Conv_632[FLOAT, 24x24x1x1]
%onnx::Conv_635[FLOAT, 24x1x5x5]
%onnx::Conv_638[FLOAT, 32x24x1x1]
%onnx::Conv_639[FLOAT, 32]
%onnx::Conv_641[FLOAT, 192x32x1x1]
%onnx::Conv_642[FLOAT, 192]
%onnx::Conv_644[FLOAT, 192x1x3x3]
%onnx::Conv_647[FLOAT, 32x192x1x1]
%onnx::Conv_650[FLOAT, 96x32x1x1]
%onnx::Conv_653[FLOAT, 96x1x3x3]
%onnx::Conv_656[FLOAT, 32x96x1x1]
%onnx::Conv_659[FLOAT, 192x32x1x1]
%onnx::Conv_662[FLOAT, 192x1x5x5]
%onnx::Conv_665[FLOAT, 32x192x1x1]
%onnx::Conv_668[FLOAT, 32x32x1x1]
%onnx::Conv_671[FLOAT, 32x1x5x5]
%onnx::Conv_674[FLOAT, 64x32x1x1]
%onnx::Conv_675[FLOAT, 64]
%onnx::Conv_677[FLOAT, 192x64x1x1]
%onnx::Conv_680[FLOAT, 192x1x5x5]
%onnx::Conv_683[FLOAT, 64x192x1x1]
%onnx::Conv_686[FLOAT, 384x64x1x1]
%onnx::Conv_687[FLOAT, 384]
%onnx::Conv_689[FLOAT, 384x1x3x3]
%onnx::Conv_692[FLOAT, 64x384x1x1]
%onnx::Conv_695[FLOAT, 192x64x1x1]
%onnx::Conv_698[FLOAT, 192x1x3x3]
%onnx::Conv_701[FLOAT, 64x192x1x1]
%onnx::Conv_704[FLOAT, 64x32x1x1]
%onnx::Conv_707[FLOAT, 64x1x5x5]
%onnx::Conv_710[FLOAT, 112x32x1x1]
%onnx::Conv_711[FLOAT, 112]
%onnx::Conv_713[FLOAT, 112x112x1x1]
%onnx::Conv_716[FLOAT, 112x1x5x5]
%onnx::Conv_719[FLOAT, 112x112x1x1]
%onnx::Conv_722[FLOAT, 336x112x1x1]
%onnx::Conv_723[FLOAT, 336]
%onnx::Conv_725[FLOAT, 336x1x5x5]
%onnx::Conv_728[FLOAT, 112x336x1x1]
%onnx::Conv_731[FLOAT, 336x112x1x1]
%onnx::Conv_734[FLOAT, 336x1x5x5]
%onnx::Conv_737[FLOAT, 112x336x1x1]
%onnx::Conv_740[FLOAT, 672x112x1x1]
%onnx::Conv_741[FLOAT, 672]
%onnx::Conv_743[FLOAT, 672x1x5x5]
%onnx::Conv_746[FLOAT, 184x672x1x1]
%onnx::Conv_747[FLOAT, 184]
%onnx::Conv_749[FLOAT, 552x184x1x1]
%onnx::Conv_750[FLOAT, 552]
%onnx::Conv_752[FLOAT, 552x1x5x5]
%onnx::Conv_755[FLOAT, 184x552x1x1]
%onnx::Conv_758[FLOAT, 184x184x1x1]
%onnx::Conv_761[FLOAT, 184x1x3x3]
%onnx::Conv_764[FLOAT, 352x184x1x1]
%onnx::Conv_765[FLOAT, 352]
%onnx::Conv_767[FLOAT, 1504x352x1x1]
%onnx::Conv_768[FLOAT, 1504]
) {
%onnx::Conv_762 = Identity(%onnx::Conv_747)
%onnx::Conv_759 = Identity(%onnx::Conv_747)
%onnx::Conv_756 = Identity(%onnx::Conv_747)
%onnx::Conv_753 = Identity(%onnx::Conv_750)
%onnx::Conv_744 = Identity(%onnx::Conv_741)
%onnx::Conv_738 = Identity(%onnx::Conv_711)
%onnx::Conv_735 = Identity(%onnx::Conv_723)
%onnx::Conv_732 = Identity(%onnx::Conv_723)
%onnx::Conv_729 = Identity(%onnx::Conv_711)
%onnx::Conv_726 = Identity(%onnx::Conv_723)
%onnx::Conv_720 = Identity(%onnx::Conv_711)
%onnx::Conv_717 = Identity(%onnx::Conv_711)
%onnx::Conv_714 = Identity(%onnx::Conv_711)
%onnx::Conv_708 = Identity(%onnx::Conv_675)
%onnx::Conv_705 = Identity(%onnx::Conv_675)
%onnx::Conv_702 = Identity(%onnx::Conv_675)
%onnx::Conv_699 = Identity(%onnx::Conv_642)
%onnx::Conv_696 = Identity(%onnx::Conv_642)
%onnx::Conv_693 = Identity(%onnx::Conv_675)
%onnx::Conv_690 = Identity(%onnx::Conv_687)
%onnx::Conv_684 = Identity(%onnx::Conv_675)
%onnx::Conv_681 = Identity(%onnx::Conv_642)
%onnx::Conv_678 = Identity(%onnx::Conv_642)
%onnx::Conv_672 = Identity(%onnx::Conv_639)
%onnx::Conv_669 = Identity(%onnx::Conv_639)
%onnx::Conv_666 = Identity(%onnx::Conv_639)
%onnx::Conv_663 = Identity(%onnx::Conv_642)
%onnx::Conv_660 = Identity(%onnx::Conv_642)
%onnx::Conv_657 = Identity(%onnx::Conv_639)
%onnx::Conv_654 = Identity(%onnx::Conv_588)
%onnx::Conv_651 = Identity(%onnx::Conv_588)
%onnx::Conv_648 = Identity(%onnx::Conv_639)
%onnx::Conv_645 = Identity(%onnx::Conv_642)
%onnx::Conv_636 = Identity(%onnx::Conv_603)
%onnx::Conv_633 = Identity(%onnx::Conv_603)
%onnx::Conv_630 = Identity(%onnx::Conv_603)
%onnx::Conv_627 = Identity(%onnx::Conv_615)
%onnx::Conv_624 = Identity(%onnx::Conv_615)
%onnx::Conv_621 = Identity(%onnx::Conv_603)
%onnx::Conv_618 = Identity(%onnx::Conv_615)
%onnx::Conv_612 = Identity(%onnx::Conv_603)
%onnx::Conv_609 = Identity(%onnx::Conv_603)
%onnx::Conv_606 = Identity(%onnx::Conv_603)
%onnx::Conv_600 = Identity(%onnx::Conv_585)
%onnx::Conv_597 = Identity(%onnx::Conv_585)
%onnx::Conv_594 = Identity(%onnx::Conv_585)
%onnx::Conv_591 = Identity(%onnx::Conv_588)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_584, %onnx::Conv_585)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_587, %onnx::Conv_588)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_590, %onnx::Conv_591)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_593, %onnx::Conv_594)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_596, %onnx::Conv_597)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_599, %onnx::Conv_600)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_602, %onnx::Conv_603)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_605, %onnx::Conv_606)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_608, %onnx::Conv_609)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_611, %onnx::Conv_612)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_614, %onnx::Conv_615)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_617, %onnx::Conv_618)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_620, %onnx::Conv_621)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_623, %onnx::Conv_624)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%582 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %582
}
|
val_accuracy
| 0
| 72,222,080
| 1,605,876
|
{'zcp_synflow': 81.02837135408502, 'zcp_zen': 69.22212219238281, 'zcp_epe_nas': 12.087941830226807, 'zcp_fisher': 0.15561823546886444, 'zcp_flops': 72222080.0, 'zcp_grad_norm': 24.090192794799805, 'zcp_grasp': -0.10617828369140625, 'zcp_jacov': -16.07185676657113, 'zcp_l2_norm': 616.7440185546875, 'zcp_nwot': 214.10373115465092, 'zcp_params': 1605876.0, 'zcp_plain': 0.003044561482965946, 'zcp_snip': 42.843467712402344, 'lat_1080ti_1': 0.45604031880820567, 'lat_1080ti_32': 0.47421300493555824, 'lat_1080ti_64': 0.43368286111428145, 'lat_2080ti_1': 0.5069226251638469, 'lat_2080ti_32': 0.451465319651643, 'lat_2080ti_64': 0.43931039305616293, 'lat_essential_ph_1': 0.41509433962264153, 'lat_eyeriss': 0.4923074101919941, 'lat_fpga': 0.4873885552004273, 'lat_gold_6226': 0.34599257601148675, 'lat_gold_6240': 0.4620684356514438, 'lat_pixel2': 0.3695652173913043, 'lat_pixel3': 0.4758465468917547, 'lat_raspi4': 0.4172611572238567, 'lat_samsung_a50': 0.17894736842105263, 'lat_samsung_s7': 0.14960629921259844, 'lat_silver_4114': 0.5151705278083161, 'lat_silver_4210r': 0.505228466727955, 'lat_titan_rtx_1': 0.4986489016177166, 'lat_titan_rtx_32': 0.43603199913009594, 'lat_titan_rtx_64': 0.44499112104226274, 'lat_titanx_1': 0.2610788523850194, 'lat_titanx_32': 0.4575372065951395, 'lat_titanx_64': 0.41135876919738207, 'lat_titanxp_1': 0.47437679042036235, 'lat_titanxp_32': 0.44794796531495573, 'lat_titanxp_64': 0.43145061065357354}
| |
FBNet_2191
|
FBNet
|
2191
|
2191
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_668[FLOAT, 16x3x3x3]
%onnx::Conv_669[FLOAT, 16]
%onnx::Conv_671[FLOAT, 48x16x1x1]
%onnx::Conv_672[FLOAT, 48]
%onnx::Conv_674[FLOAT, 48x1x5x5]
%onnx::Conv_677[FLOAT, 16x48x1x1]
%onnx::Conv_680[FLOAT, 16x16x1x1]
%onnx::Conv_683[FLOAT, 16x1x5x5]
%onnx::Conv_686[FLOAT, 24x16x1x1]
%onnx::Conv_687[FLOAT, 24]
%onnx::Conv_689[FLOAT, 72x24x1x1]
%onnx::Conv_690[FLOAT, 72]
%onnx::Conv_692[FLOAT, 72x1x5x5]
%onnx::Conv_695[FLOAT, 24x72x1x1]
%onnx::Conv_698[FLOAT, 72x24x1x1]
%onnx::Conv_701[FLOAT, 72x1x3x3]
%onnx::Conv_704[FLOAT, 24x72x1x1]
%onnx::Conv_707[FLOAT, 24x24x1x1]
%onnx::Conv_710[FLOAT, 24x1x3x3]
%onnx::Conv_713[FLOAT, 24x24x1x1]
%onnx::Conv_716[FLOAT, 24x12x1x1]
%onnx::Conv_719[FLOAT, 24x1x3x3]
%onnx::Conv_722[FLOAT, 32x12x1x1]
%onnx::Conv_723[FLOAT, 32]
%onnx::Conv_725[FLOAT, 32x16x1x1]
%onnx::Conv_728[FLOAT, 32x1x3x3]
%onnx::Conv_731[FLOAT, 32x16x1x1]
%onnx::Conv_734[FLOAT, 192x32x1x1]
%onnx::Conv_735[FLOAT, 192]
%onnx::Conv_737[FLOAT, 192x1x3x3]
%onnx::Conv_740[FLOAT, 32x192x1x1]
%onnx::Conv_743[FLOAT, 96x32x1x1]
%onnx::Conv_744[FLOAT, 96]
%onnx::Conv_746[FLOAT, 96x1x3x3]
%onnx::Conv_749[FLOAT, 32x96x1x1]
%onnx::Conv_752[FLOAT, 96x32x1x1]
%onnx::Conv_755[FLOAT, 96x1x5x5]
%onnx::Conv_758[FLOAT, 64x96x1x1]
%onnx::Conv_759[FLOAT, 64]
%onnx::Conv_761[FLOAT, 64x32x1x1]
%onnx::Conv_764[FLOAT, 64x1x3x3]
%onnx::Conv_767[FLOAT, 64x32x1x1]
%onnx::Conv_770[FLOAT, 64x64x1x1]
%onnx::Conv_773[FLOAT, 64x1x3x3]
%onnx::Conv_776[FLOAT, 64x64x1x1]
%onnx::Conv_779[FLOAT, 64x64x1x1]
%onnx::Conv_782[FLOAT, 64x1x5x5]
%onnx::Conv_785[FLOAT, 112x64x1x1]
%onnx::Conv_786[FLOAT, 112]
%onnx::Conv_788[FLOAT, 112x56x1x1]
%onnx::Conv_791[FLOAT, 112x1x3x3]
%onnx::Conv_794[FLOAT, 112x56x1x1]
%onnx::Conv_797[FLOAT, 336x112x1x1]
%onnx::Conv_798[FLOAT, 336]
%onnx::Conv_800[FLOAT, 336x1x3x3]
%onnx::Conv_803[FLOAT, 112x336x1x1]
%onnx::Conv_806[FLOAT, 112x112x1x1]
%onnx::Conv_809[FLOAT, 112x1x5x5]
%onnx::Conv_812[FLOAT, 112x112x1x1]
%onnx::Conv_815[FLOAT, 672x112x1x1]
%onnx::Conv_816[FLOAT, 672]
%onnx::Conv_818[FLOAT, 672x1x3x3]
%onnx::Conv_821[FLOAT, 184x672x1x1]
%onnx::Conv_822[FLOAT, 184]
%onnx::Conv_824[FLOAT, 552x184x1x1]
%onnx::Conv_825[FLOAT, 552]
%onnx::Conv_827[FLOAT, 552x1x3x3]
%onnx::Conv_830[FLOAT, 184x552x1x1]
%onnx::Conv_833[FLOAT, 1104x184x1x1]
%onnx::Conv_834[FLOAT, 1104]
%onnx::Conv_836[FLOAT, 1104x1x3x3]
%onnx::Conv_839[FLOAT, 184x1104x1x1]
%onnx::Conv_842[FLOAT, 552x184x1x1]
%onnx::Conv_845[FLOAT, 552x1x3x3]
%onnx::Conv_848[FLOAT, 184x552x1x1]
%onnx::Conv_851[FLOAT, 184x184x1x1]
%onnx::Conv_854[FLOAT, 184x1x3x3]
%onnx::Conv_857[FLOAT, 352x184x1x1]
%onnx::Conv_858[FLOAT, 352]
%onnx::Conv_860[FLOAT, 1504x352x1x1]
%onnx::Conv_861[FLOAT, 1504]
) {
%onnx::Conv_855 = Identity(%onnx::Conv_822)
%onnx::Conv_852 = Identity(%onnx::Conv_822)
%onnx::Conv_849 = Identity(%onnx::Conv_822)
%onnx::Conv_846 = Identity(%onnx::Conv_825)
%onnx::Conv_843 = Identity(%onnx::Conv_825)
%onnx::Conv_840 = Identity(%onnx::Conv_822)
%onnx::Conv_837 = Identity(%onnx::Conv_834)
%onnx::Conv_831 = Identity(%onnx::Conv_822)
%onnx::Conv_828 = Identity(%onnx::Conv_825)
%onnx::Conv_819 = Identity(%onnx::Conv_816)
%onnx::Conv_813 = Identity(%onnx::Conv_786)
%onnx::Conv_810 = Identity(%onnx::Conv_786)
%onnx::Conv_807 = Identity(%onnx::Conv_786)
%onnx::Conv_804 = Identity(%onnx::Conv_786)
%onnx::Conv_801 = Identity(%onnx::Conv_798)
%onnx::Conv_795 = Identity(%onnx::Conv_786)
%onnx::Conv_792 = Identity(%onnx::Conv_786)
%onnx::Conv_789 = Identity(%onnx::Conv_786)
%onnx::Conv_783 = Identity(%onnx::Conv_759)
%onnx::Conv_780 = Identity(%onnx::Conv_759)
%onnx::Conv_777 = Identity(%onnx::Conv_759)
%onnx::Conv_774 = Identity(%onnx::Conv_759)
%onnx::Conv_771 = Identity(%onnx::Conv_759)
%onnx::Conv_768 = Identity(%onnx::Conv_759)
%onnx::Conv_765 = Identity(%onnx::Conv_759)
%onnx::Conv_762 = Identity(%onnx::Conv_759)
%onnx::Conv_756 = Identity(%onnx::Conv_744)
%onnx::Conv_753 = Identity(%onnx::Conv_744)
%onnx::Conv_750 = Identity(%onnx::Conv_723)
%onnx::Conv_747 = Identity(%onnx::Conv_744)
%onnx::Conv_741 = Identity(%onnx::Conv_723)
%onnx::Conv_738 = Identity(%onnx::Conv_735)
%onnx::Conv_732 = Identity(%onnx::Conv_723)
%onnx::Conv_729 = Identity(%onnx::Conv_723)
%onnx::Conv_726 = Identity(%onnx::Conv_723)
%onnx::Conv_720 = Identity(%onnx::Conv_687)
%onnx::Conv_717 = Identity(%onnx::Conv_687)
%onnx::Conv_714 = Identity(%onnx::Conv_687)
%onnx::Conv_711 = Identity(%onnx::Conv_687)
%onnx::Conv_708 = Identity(%onnx::Conv_687)
%onnx::Conv_705 = Identity(%onnx::Conv_687)
%onnx::Conv_702 = Identity(%onnx::Conv_690)
%onnx::Conv_699 = Identity(%onnx::Conv_690)
%onnx::Conv_696 = Identity(%onnx::Conv_687)
%onnx::Conv_693 = Identity(%onnx::Conv_690)
%onnx::Conv_684 = Identity(%onnx::Conv_669)
%onnx::Conv_681 = Identity(%onnx::Conv_669)
%onnx::Conv_678 = Identity(%onnx::Conv_669)
%onnx::Conv_675 = Identity(%onnx::Conv_672)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_668, %onnx::Conv_669)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_803, %onnx::Conv_804)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_806, %onnx::Conv_807)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_809, %onnx::Conv_810)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_812, %onnx::Conv_813)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_815, %onnx::Conv_816)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_818, %onnx::Conv_819)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_821, %onnx::Conv_822)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_824, %onnx::Conv_825)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_827, %onnx::Conv_828)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_830, %onnx::Conv_831)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_833, %onnx::Conv_834)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_836, %onnx::Conv_837)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_839, %onnx::Conv_840)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_842, %onnx::Conv_843)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_845, %onnx::Conv_846)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_848, %onnx::Conv_849)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_851, %onnx::Conv_852)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_854, %onnx::Conv_855)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_857, %onnx::Conv_858)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_860, %onnx::Conv_861)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%666 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %666
}
|
val_accuracy
| 0
| 65,416,320
| 2,038,300
|
{'zcp_synflow': 79.54302704310585, 'zcp_zen': 69.43112182617188, 'zcp_epe_nas': 6.073424049763659, 'zcp_fisher': 0.2219117283821106, 'zcp_flops': 65416320.0, 'zcp_grad_norm': 26.052425384521484, 'zcp_grasp': 0.19143295288085938, 'zcp_jacov': -16.050184591603294, 'zcp_l2_norm': 643.8694458007812, 'zcp_nwot': 210.96055412560906, 'zcp_params': 2038300.0, 'zcp_plain': -0.0033001606352627277, 'zcp_snip': 45.0604362487793, 'lat_1080ti_1': 0.6113844059765816, 'lat_1080ti_32': 0.6277615785115764, 'lat_1080ti_64': 0.42100989166545094, 'lat_2080ti_1': 0.696130833110325, 'lat_2080ti_32': 0.561513338455999, 'lat_2080ti_64': 0.4358777590395303, 'lat_essential_ph_1': 0.4339622641509434, 'lat_eyeriss': 0.4077164285845267, 'lat_fpga': 0.4273048417261202, 'lat_gold_6226': 0.34957072742616396, 'lat_gold_6240': 0.5706763848336618, 'lat_pixel2': 0.30434782608695654, 'lat_pixel3': 0.3710892028073602, 'lat_raspi4': 0.37393971150043726, 'lat_samsung_a50': 0.18947368421052632, 'lat_samsung_s7': 0.1889763779527559, 'lat_silver_4114': 0.6342619469356079, 'lat_silver_4210r': 0.5804100123477764, 'lat_titan_rtx_1': 0.6653227448506851, 'lat_titan_rtx_32': 0.5609611266434807, 'lat_titan_rtx_64': 0.47844525942062843, 'lat_titanx_1': 0.3581835931811682, 'lat_titanx_32': 0.5160073835473475, 'lat_titanx_64': 0.3859799437538941, 'lat_titanxp_1': 0.6495314294463207, 'lat_titanxp_32': 0.5821943272217495, 'lat_titanxp_64': 0.4408202414724729}
| |
FBNet_754
|
FBNet
|
754
|
754
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_658[FLOAT, 16x3x3x3]
%onnx::Conv_659[FLOAT, 16]
%onnx::Conv_661[FLOAT, 16x16x1x1]
%onnx::Conv_664[FLOAT, 16x1x5x5]
%onnx::Conv_667[FLOAT, 16x16x1x1]
%onnx::Conv_670[FLOAT, 16x16x1x1]
%onnx::Conv_673[FLOAT, 16x1x3x3]
%onnx::Conv_676[FLOAT, 24x16x1x1]
%onnx::Conv_677[FLOAT, 24]
%onnx::Conv_679[FLOAT, 144x24x1x1]
%onnx::Conv_680[FLOAT, 144]
%onnx::Conv_682[FLOAT, 144x1x5x5]
%onnx::Conv_685[FLOAT, 24x144x1x1]
%onnx::Conv_688[FLOAT, 72x24x1x1]
%onnx::Conv_689[FLOAT, 72]
%onnx::Conv_691[FLOAT, 72x1x3x3]
%onnx::Conv_694[FLOAT, 24x72x1x1]
%onnx::Conv_697[FLOAT, 24x24x1x1]
%onnx::Conv_700[FLOAT, 24x1x3x3]
%onnx::Conv_703[FLOAT, 24x24x1x1]
%onnx::Conv_706[FLOAT, 24x24x1x1]
%onnx::Conv_709[FLOAT, 24x1x3x3]
%onnx::Conv_712[FLOAT, 32x24x1x1]
%onnx::Conv_713[FLOAT, 32]
%onnx::Conv_715[FLOAT, 192x32x1x1]
%onnx::Conv_716[FLOAT, 192]
%onnx::Conv_718[FLOAT, 192x1x3x3]
%onnx::Conv_721[FLOAT, 32x192x1x1]
%onnx::Conv_724[FLOAT, 32x32x1x1]
%onnx::Conv_727[FLOAT, 32x1x3x3]
%onnx::Conv_730[FLOAT, 32x32x1x1]
%onnx::Conv_733[FLOAT, 32x16x1x1]
%onnx::Conv_736[FLOAT, 32x1x3x3]
%onnx::Conv_739[FLOAT, 32x16x1x1]
%onnx::Conv_742[FLOAT, 32x16x1x1]
%onnx::Conv_745[FLOAT, 32x1x3x3]
%onnx::Conv_748[FLOAT, 64x16x1x1]
%onnx::Conv_749[FLOAT, 64]
%onnx::Conv_751[FLOAT, 192x64x1x1]
%onnx::Conv_754[FLOAT, 192x1x5x5]
%onnx::Conv_757[FLOAT, 64x192x1x1]
%onnx::Conv_760[FLOAT, 192x64x1x1]
%onnx::Conv_763[FLOAT, 192x1x5x5]
%onnx::Conv_766[FLOAT, 64x192x1x1]
%onnx::Conv_769[FLOAT, 192x64x1x1]
%onnx::Conv_772[FLOAT, 192x1x3x3]
%onnx::Conv_775[FLOAT, 112x192x1x1]
%onnx::Conv_776[FLOAT, 112]
%onnx::Conv_778[FLOAT, 336x112x1x1]
%onnx::Conv_779[FLOAT, 336]
%onnx::Conv_781[FLOAT, 336x1x5x5]
%onnx::Conv_784[FLOAT, 112x336x1x1]
%onnx::Conv_787[FLOAT, 112x56x1x1]
%onnx::Conv_790[FLOAT, 112x1x3x3]
%onnx::Conv_793[FLOAT, 112x56x1x1]
%onnx::Conv_796[FLOAT, 112x112x1x1]
%onnx::Conv_799[FLOAT, 112x1x5x5]
%onnx::Conv_802[FLOAT, 184x112x1x1]
%onnx::Conv_803[FLOAT, 184]
%onnx::Conv_805[FLOAT, 1104x184x1x1]
%onnx::Conv_806[FLOAT, 1104]
%onnx::Conv_808[FLOAT, 1104x1x3x3]
%onnx::Conv_811[FLOAT, 184x1104x1x1]
%onnx::Conv_814[FLOAT, 184x92x1x1]
%onnx::Conv_817[FLOAT, 184x1x3x3]
%onnx::Conv_820[FLOAT, 184x92x1x1]
%onnx::Conv_823[FLOAT, 184x92x1x1]
%onnx::Conv_826[FLOAT, 184x1x3x3]
%onnx::Conv_829[FLOAT, 184x92x1x1]
%onnx::Conv_832[FLOAT, 1104x184x1x1]
%onnx::Conv_835[FLOAT, 1104x1x5x5]
%onnx::Conv_838[FLOAT, 352x1104x1x1]
%onnx::Conv_839[FLOAT, 352]
%onnx::Conv_841[FLOAT, 1504x352x1x1]
%onnx::Conv_842[FLOAT, 1504]
) {
%onnx::Conv_836 = Identity(%onnx::Conv_806)
%onnx::Conv_833 = Identity(%onnx::Conv_806)
%onnx::Conv_830 = Identity(%onnx::Conv_803)
%onnx::Conv_827 = Identity(%onnx::Conv_803)
%onnx::Conv_824 = Identity(%onnx::Conv_803)
%onnx::Conv_821 = Identity(%onnx::Conv_803)
%onnx::Conv_818 = Identity(%onnx::Conv_803)
%onnx::Conv_815 = Identity(%onnx::Conv_803)
%onnx::Conv_812 = Identity(%onnx::Conv_803)
%onnx::Conv_809 = Identity(%onnx::Conv_806)
%onnx::Conv_800 = Identity(%onnx::Conv_776)
%onnx::Conv_797 = Identity(%onnx::Conv_776)
%onnx::Conv_794 = Identity(%onnx::Conv_776)
%onnx::Conv_791 = Identity(%onnx::Conv_776)
%onnx::Conv_788 = Identity(%onnx::Conv_776)
%onnx::Conv_785 = Identity(%onnx::Conv_776)
%onnx::Conv_782 = Identity(%onnx::Conv_779)
%onnx::Conv_773 = Identity(%onnx::Conv_716)
%onnx::Conv_770 = Identity(%onnx::Conv_716)
%onnx::Conv_767 = Identity(%onnx::Conv_749)
%onnx::Conv_764 = Identity(%onnx::Conv_716)
%onnx::Conv_761 = Identity(%onnx::Conv_716)
%onnx::Conv_758 = Identity(%onnx::Conv_749)
%onnx::Conv_755 = Identity(%onnx::Conv_716)
%onnx::Conv_752 = Identity(%onnx::Conv_716)
%onnx::Conv_746 = Identity(%onnx::Conv_713)
%onnx::Conv_743 = Identity(%onnx::Conv_713)
%onnx::Conv_740 = Identity(%onnx::Conv_713)
%onnx::Conv_737 = Identity(%onnx::Conv_713)
%onnx::Conv_734 = Identity(%onnx::Conv_713)
%onnx::Conv_731 = Identity(%onnx::Conv_713)
%onnx::Conv_728 = Identity(%onnx::Conv_713)
%onnx::Conv_725 = Identity(%onnx::Conv_713)
%onnx::Conv_722 = Identity(%onnx::Conv_713)
%onnx::Conv_719 = Identity(%onnx::Conv_716)
%onnx::Conv_710 = Identity(%onnx::Conv_677)
%onnx::Conv_707 = Identity(%onnx::Conv_677)
%onnx::Conv_704 = Identity(%onnx::Conv_677)
%onnx::Conv_701 = Identity(%onnx::Conv_677)
%onnx::Conv_698 = Identity(%onnx::Conv_677)
%onnx::Conv_695 = Identity(%onnx::Conv_677)
%onnx::Conv_692 = Identity(%onnx::Conv_689)
%onnx::Conv_686 = Identity(%onnx::Conv_677)
%onnx::Conv_683 = Identity(%onnx::Conv_680)
%onnx::Conv_674 = Identity(%onnx::Conv_659)
%onnx::Conv_671 = Identity(%onnx::Conv_659)
%onnx::Conv_668 = Identity(%onnx::Conv_659)
%onnx::Conv_665 = Identity(%onnx::Conv_659)
%onnx::Conv_662 = Identity(%onnx::Conv_659)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_658, %onnx::Conv_659)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_823, %onnx::Conv_824)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_826, %onnx::Conv_827)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_829, %onnx::Conv_830)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_832, %onnx::Conv_833)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_835, %onnx::Conv_836)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_838, %onnx::Conv_839)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_841, %onnx::Conv_842)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%656 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %656
}
|
val_accuracy
| 0
| 66,790,784
| 2,077,020
|
{'zcp_synflow': 73.7655891855748, 'zcp_zen': 64.51087951660156, 'zcp_epe_nas': 10.80797724823031, 'zcp_fisher': 0.09862842410802841, 'zcp_flops': 66790784.0, 'zcp_grad_norm': 21.0572509765625, 'zcp_grasp': -0.14227676391601562, 'zcp_jacov': -16.069554168072933, 'zcp_l2_norm': 588.4036254882812, 'zcp_nwot': 211.5191968473724, 'zcp_params': 2077020.0, 'zcp_plain': 0.009114754386246204, 'zcp_snip': 38.75285720825195, 'lat_1080ti_1': 0.6118130483161547, 'lat_1080ti_32': 0.6073860363584411, 'lat_1080ti_64': 0.4664609699241906, 'lat_2080ti_1': 0.6064649516245633, 'lat_2080ti_32': 0.5888021168362432, 'lat_2080ti_64': 0.5139850550551179, 'lat_essential_ph_1': 0.3018867924528302, 'lat_eyeriss': 0.4294830652999102, 'lat_fpga': 0.46982455793150535, 'lat_gold_6226': 0.3336209618630341, 'lat_gold_6240': 0.544022500149207, 'lat_pixel2': 0.34782608695652173, 'lat_pixel3': 0.43324491534999643, 'lat_raspi4': 0.5264081514808802, 'lat_samsung_a50': 0.18947368421052632, 'lat_samsung_s7': 0.2204724409448819, 'lat_silver_4114': 0.5059783709020318, 'lat_silver_4210r': 0.5250716358422801, 'lat_titan_rtx_1': 0.5853249396997681, 'lat_titan_rtx_32': 0.5736324524665922, 'lat_titan_rtx_64': 0.5154800491990182, 'lat_titanx_1': 0.31240666482812884, 'lat_titanx_32': 0.544934119433525, 'lat_titanx_64': 0.43216700602896785, 'lat_titanxp_1': 0.5432752025689045, 'lat_titanxp_32': 0.5616465097981516, 'lat_titanxp_64': 0.47711996731404654}
| |
FBNet_3723
|
FBNet
|
3723
|
3723
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_640[FLOAT, 16x3x3x3]
%onnx::Conv_641[FLOAT, 16]
%onnx::Conv_643[FLOAT, 16x8x1x1]
%onnx::Conv_646[FLOAT, 16x1x5x5]
%onnx::Conv_649[FLOAT, 16x8x1x1]
%onnx::Conv_652[FLOAT, 16x8x1x1]
%onnx::Conv_655[FLOAT, 16x1x3x3]
%onnx::Conv_658[FLOAT, 24x8x1x1]
%onnx::Conv_659[FLOAT, 24]
%onnx::Conv_661[FLOAT, 144x24x1x1]
%onnx::Conv_662[FLOAT, 144]
%onnx::Conv_664[FLOAT, 144x1x5x5]
%onnx::Conv_667[FLOAT, 24x144x1x1]
%onnx::Conv_670[FLOAT, 72x24x1x1]
%onnx::Conv_671[FLOAT, 72]
%onnx::Conv_673[FLOAT, 72x1x5x5]
%onnx::Conv_676[FLOAT, 24x72x1x1]
%onnx::Conv_679[FLOAT, 144x24x1x1]
%onnx::Conv_682[FLOAT, 144x1x5x5]
%onnx::Conv_685[FLOAT, 24x144x1x1]
%onnx::Conv_688[FLOAT, 72x24x1x1]
%onnx::Conv_691[FLOAT, 72x1x3x3]
%onnx::Conv_694[FLOAT, 32x72x1x1]
%onnx::Conv_695[FLOAT, 32]
%onnx::Conv_697[FLOAT, 32x32x1x1]
%onnx::Conv_700[FLOAT, 32x1x3x3]
%onnx::Conv_703[FLOAT, 32x32x1x1]
%onnx::Conv_706[FLOAT, 192x32x1x1]
%onnx::Conv_707[FLOAT, 192]
%onnx::Conv_709[FLOAT, 192x1x5x5]
%onnx::Conv_712[FLOAT, 32x192x1x1]
%onnx::Conv_715[FLOAT, 96x32x1x1]
%onnx::Conv_716[FLOAT, 96]
%onnx::Conv_718[FLOAT, 96x1x3x3]
%onnx::Conv_721[FLOAT, 64x96x1x1]
%onnx::Conv_722[FLOAT, 64]
%onnx::Conv_724[FLOAT, 64x32x1x1]
%onnx::Conv_727[FLOAT, 64x1x3x3]
%onnx::Conv_730[FLOAT, 64x32x1x1]
%onnx::Conv_733[FLOAT, 384x64x1x1]
%onnx::Conv_734[FLOAT, 384]
%onnx::Conv_736[FLOAT, 384x1x3x3]
%onnx::Conv_739[FLOAT, 64x384x1x1]
%onnx::Conv_742[FLOAT, 192x64x1x1]
%onnx::Conv_745[FLOAT, 192x1x3x3]
%onnx::Conv_748[FLOAT, 64x192x1x1]
%onnx::Conv_751[FLOAT, 384x64x1x1]
%onnx::Conv_754[FLOAT, 384x1x5x5]
%onnx::Conv_757[FLOAT, 112x384x1x1]
%onnx::Conv_758[FLOAT, 112]
%onnx::Conv_760[FLOAT, 112x56x1x1]
%onnx::Conv_763[FLOAT, 112x1x5x5]
%onnx::Conv_766[FLOAT, 112x56x1x1]
%onnx::Conv_769[FLOAT, 336x112x1x1]
%onnx::Conv_770[FLOAT, 336]
%onnx::Conv_772[FLOAT, 336x1x3x3]
%onnx::Conv_775[FLOAT, 112x336x1x1]
%onnx::Conv_778[FLOAT, 336x112x1x1]
%onnx::Conv_781[FLOAT, 336x1x3x3]
%onnx::Conv_784[FLOAT, 184x336x1x1]
%onnx::Conv_785[FLOAT, 184]
%onnx::Conv_787[FLOAT, 552x184x1x1]
%onnx::Conv_788[FLOAT, 552]
%onnx::Conv_790[FLOAT, 552x1x5x5]
%onnx::Conv_793[FLOAT, 184x552x1x1]
%onnx::Conv_796[FLOAT, 184x184x1x1]
%onnx::Conv_799[FLOAT, 184x1x3x3]
%onnx::Conv_802[FLOAT, 184x184x1x1]
%onnx::Conv_805[FLOAT, 552x184x1x1]
%onnx::Conv_808[FLOAT, 552x1x5x5]
%onnx::Conv_811[FLOAT, 184x552x1x1]
%onnx::Conv_814[FLOAT, 552x184x1x1]
%onnx::Conv_817[FLOAT, 552x1x5x5]
%onnx::Conv_820[FLOAT, 352x552x1x1]
%onnx::Conv_821[FLOAT, 352]
%onnx::Conv_823[FLOAT, 1504x352x1x1]
%onnx::Conv_824[FLOAT, 1504]
) {
%onnx::Conv_818 = Identity(%onnx::Conv_788)
%onnx::Conv_815 = Identity(%onnx::Conv_788)
%onnx::Conv_812 = Identity(%onnx::Conv_785)
%onnx::Conv_809 = Identity(%onnx::Conv_788)
%onnx::Conv_806 = Identity(%onnx::Conv_788)
%onnx::Conv_803 = Identity(%onnx::Conv_785)
%onnx::Conv_800 = Identity(%onnx::Conv_785)
%onnx::Conv_797 = Identity(%onnx::Conv_785)
%onnx::Conv_794 = Identity(%onnx::Conv_785)
%onnx::Conv_791 = Identity(%onnx::Conv_788)
%onnx::Conv_782 = Identity(%onnx::Conv_770)
%onnx::Conv_779 = Identity(%onnx::Conv_770)
%onnx::Conv_776 = Identity(%onnx::Conv_758)
%onnx::Conv_773 = Identity(%onnx::Conv_770)
%onnx::Conv_767 = Identity(%onnx::Conv_758)
%onnx::Conv_764 = Identity(%onnx::Conv_758)
%onnx::Conv_761 = Identity(%onnx::Conv_758)
%onnx::Conv_755 = Identity(%onnx::Conv_734)
%onnx::Conv_752 = Identity(%onnx::Conv_734)
%onnx::Conv_749 = Identity(%onnx::Conv_722)
%onnx::Conv_746 = Identity(%onnx::Conv_707)
%onnx::Conv_743 = Identity(%onnx::Conv_707)
%onnx::Conv_740 = Identity(%onnx::Conv_722)
%onnx::Conv_737 = Identity(%onnx::Conv_734)
%onnx::Conv_731 = Identity(%onnx::Conv_722)
%onnx::Conv_728 = Identity(%onnx::Conv_722)
%onnx::Conv_725 = Identity(%onnx::Conv_722)
%onnx::Conv_719 = Identity(%onnx::Conv_716)
%onnx::Conv_713 = Identity(%onnx::Conv_695)
%onnx::Conv_710 = Identity(%onnx::Conv_707)
%onnx::Conv_704 = Identity(%onnx::Conv_695)
%onnx::Conv_701 = Identity(%onnx::Conv_695)
%onnx::Conv_698 = Identity(%onnx::Conv_695)
%onnx::Conv_692 = Identity(%onnx::Conv_671)
%onnx::Conv_689 = Identity(%onnx::Conv_671)
%onnx::Conv_686 = Identity(%onnx::Conv_659)
%onnx::Conv_683 = Identity(%onnx::Conv_662)
%onnx::Conv_680 = Identity(%onnx::Conv_662)
%onnx::Conv_677 = Identity(%onnx::Conv_659)
%onnx::Conv_674 = Identity(%onnx::Conv_671)
%onnx::Conv_668 = Identity(%onnx::Conv_659)
%onnx::Conv_665 = Identity(%onnx::Conv_662)
%onnx::Conv_656 = Identity(%onnx::Conv_641)
%onnx::Conv_653 = Identity(%onnx::Conv_641)
%onnx::Conv_650 = Identity(%onnx::Conv_641)
%onnx::Conv_647 = Identity(%onnx::Conv_641)
%onnx::Conv_644 = Identity(%onnx::Conv_641)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_640, %onnx::Conv_641)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_643, %onnx::Conv_644)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_646, %onnx::Conv_647)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_649, %onnx::Conv_650)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_652, %onnx::Conv_653)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_655, %onnx::Conv_656)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_658, %onnx::Conv_659)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_823, %onnx::Conv_824)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%638 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %638
}
|
val_accuracy
| 0
| 82,363,776
| 1,936,260
|
{'zcp_synflow': 76.3389232314857, 'zcp_zen': 69.33539581298828, 'zcp_epe_nas': 16.715390479803077, 'zcp_fisher': 0.10332515835762024, 'zcp_flops': 82363776.0, 'zcp_grad_norm': 22.68507957458496, 'zcp_grasp': 0.11730766296386719, 'zcp_jacov': -16.060824162965616, 'zcp_l2_norm': 648.1148071289062, 'zcp_nwot': 216.71897907007607, 'zcp_params': 1936260.0, 'zcp_plain': -0.003911977633833885, 'zcp_snip': 42.29576110839844, 'lat_1080ti_1': 0.5376026118006195, 'lat_1080ti_32': 0.7278898775583474, 'lat_1080ti_64': 0.6893862758455616, 'lat_2080ti_1': 0.6020849355819419, 'lat_2080ti_32': 0.696934001328904, 'lat_2080ti_64': 0.6978393189001296, 'lat_essential_ph_1': 0.37735849056603776, 'lat_eyeriss': 0.6405662626299673, 'lat_fpga': 0.5819768649996537, 'lat_gold_6226': 0.42338334772430536, 'lat_gold_6240': 0.47804618866005993, 'lat_pixel2': 0.3695652173913043, 'lat_pixel3': 0.6993184917302189, 'lat_raspi4': 0.7109752340516422, 'lat_samsung_a50': 0.24210526315789474, 'lat_samsung_s7': 0.1968503937007874, 'lat_silver_4114': 0.49607540377794185, 'lat_silver_4210r': 0.49134204932097253, 'lat_titan_rtx_1': 0.5695662593941695, 'lat_titan_rtx_32': 0.6606154897372561, 'lat_titan_rtx_64': 0.7080734902504803, 'lat_titanx_1': 0.30291698703415465, 'lat_titanx_32': 0.7194895574263389, 'lat_titanx_64': 0.7091054802156677, 'lat_titanxp_1': 0.5351919101358029, 'lat_titanxp_32': 0.7010417603346561, 'lat_titanxp_64': 0.709903070912172}
| |
FBNet_697
|
FBNet
|
697
|
697
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_632[FLOAT, 16x3x3x3]
%onnx::Conv_633[FLOAT, 16]
%onnx::Conv_635[FLOAT, 96x16x1x1]
%onnx::Conv_636[FLOAT, 96]
%onnx::Conv_638[FLOAT, 96x1x5x5]
%onnx::Conv_641[FLOAT, 16x96x1x1]
%onnx::Conv_644[FLOAT, 16x8x1x1]
%onnx::Conv_647[FLOAT, 16x1x5x5]
%onnx::Conv_650[FLOAT, 24x8x1x1]
%onnx::Conv_651[FLOAT, 24]
%onnx::Conv_653[FLOAT, 72x24x1x1]
%onnx::Conv_654[FLOAT, 72]
%onnx::Conv_656[FLOAT, 72x1x3x3]
%onnx::Conv_659[FLOAT, 24x72x1x1]
%onnx::Conv_662[FLOAT, 24x24x1x1]
%onnx::Conv_665[FLOAT, 24x1x3x3]
%onnx::Conv_668[FLOAT, 24x24x1x1]
%onnx::Conv_671[FLOAT, 72x24x1x1]
%onnx::Conv_674[FLOAT, 72x1x5x5]
%onnx::Conv_677[FLOAT, 24x72x1x1]
%onnx::Conv_680[FLOAT, 24x12x1x1]
%onnx::Conv_683[FLOAT, 24x1x3x3]
%onnx::Conv_686[FLOAT, 32x12x1x1]
%onnx::Conv_687[FLOAT, 32]
%onnx::Conv_689[FLOAT, 96x32x1x1]
%onnx::Conv_692[FLOAT, 96x1x5x5]
%onnx::Conv_695[FLOAT, 32x96x1x1]
%onnx::Conv_698[FLOAT, 32x32x1x1]
%onnx::Conv_701[FLOAT, 32x1x3x3]
%onnx::Conv_704[FLOAT, 32x32x1x1]
%onnx::Conv_707[FLOAT, 32x32x1x1]
%onnx::Conv_710[FLOAT, 32x1x3x3]
%onnx::Conv_713[FLOAT, 32x32x1x1]
%onnx::Conv_716[FLOAT, 64x32x1x1]
%onnx::Conv_717[FLOAT, 64]
%onnx::Conv_719[FLOAT, 64x32x1x1]
%onnx::Conv_722[FLOAT, 64x1x3x3]
%onnx::Conv_725[FLOAT, 64x32x1x1]
%onnx::Conv_728[FLOAT, 192x64x1x1]
%onnx::Conv_729[FLOAT, 192]
%onnx::Conv_731[FLOAT, 192x1x5x5]
%onnx::Conv_734[FLOAT, 64x192x1x1]
%onnx::Conv_737[FLOAT, 192x64x1x1]
%onnx::Conv_740[FLOAT, 192x1x5x5]
%onnx::Conv_743[FLOAT, 112x192x1x1]
%onnx::Conv_744[FLOAT, 112]
%onnx::Conv_746[FLOAT, 672x112x1x1]
%onnx::Conv_747[FLOAT, 672]
%onnx::Conv_749[FLOAT, 672x1x3x3]
%onnx::Conv_752[FLOAT, 112x672x1x1]
%onnx::Conv_755[FLOAT, 112x112x1x1]
%onnx::Conv_758[FLOAT, 112x1x3x3]
%onnx::Conv_761[FLOAT, 112x112x1x1]
%onnx::Conv_764[FLOAT, 336x112x1x1]
%onnx::Conv_765[FLOAT, 336]
%onnx::Conv_767[FLOAT, 336x1x5x5]
%onnx::Conv_770[FLOAT, 112x336x1x1]
%onnx::Conv_773[FLOAT, 336x112x1x1]
%onnx::Conv_776[FLOAT, 336x1x5x5]
%onnx::Conv_779[FLOAT, 184x336x1x1]
%onnx::Conv_780[FLOAT, 184]
%onnx::Conv_782[FLOAT, 184x184x1x1]
%onnx::Conv_785[FLOAT, 184x1x3x3]
%onnx::Conv_788[FLOAT, 184x184x1x1]
%onnx::Conv_791[FLOAT, 552x184x1x1]
%onnx::Conv_792[FLOAT, 552]
%onnx::Conv_794[FLOAT, 552x1x3x3]
%onnx::Conv_797[FLOAT, 184x552x1x1]
%onnx::Conv_800[FLOAT, 184x184x1x1]
%onnx::Conv_803[FLOAT, 184x1x3x3]
%onnx::Conv_806[FLOAT, 184x184x1x1]
%onnx::Conv_809[FLOAT, 552x184x1x1]
%onnx::Conv_812[FLOAT, 552x1x5x5]
%onnx::Conv_815[FLOAT, 352x552x1x1]
%onnx::Conv_816[FLOAT, 352]
%onnx::Conv_818[FLOAT, 1504x352x1x1]
%onnx::Conv_819[FLOAT, 1504]
) {
%onnx::Conv_813 = Identity(%onnx::Conv_792)
%onnx::Conv_810 = Identity(%onnx::Conv_792)
%onnx::Conv_807 = Identity(%onnx::Conv_780)
%onnx::Conv_804 = Identity(%onnx::Conv_780)
%onnx::Conv_801 = Identity(%onnx::Conv_780)
%onnx::Conv_798 = Identity(%onnx::Conv_780)
%onnx::Conv_795 = Identity(%onnx::Conv_792)
%onnx::Conv_789 = Identity(%onnx::Conv_780)
%onnx::Conv_786 = Identity(%onnx::Conv_780)
%onnx::Conv_783 = Identity(%onnx::Conv_780)
%onnx::Conv_777 = Identity(%onnx::Conv_765)
%onnx::Conv_774 = Identity(%onnx::Conv_765)
%onnx::Conv_771 = Identity(%onnx::Conv_744)
%onnx::Conv_768 = Identity(%onnx::Conv_765)
%onnx::Conv_762 = Identity(%onnx::Conv_744)
%onnx::Conv_759 = Identity(%onnx::Conv_744)
%onnx::Conv_756 = Identity(%onnx::Conv_744)
%onnx::Conv_753 = Identity(%onnx::Conv_744)
%onnx::Conv_750 = Identity(%onnx::Conv_747)
%onnx::Conv_741 = Identity(%onnx::Conv_729)
%onnx::Conv_738 = Identity(%onnx::Conv_729)
%onnx::Conv_735 = Identity(%onnx::Conv_717)
%onnx::Conv_732 = Identity(%onnx::Conv_729)
%onnx::Conv_726 = Identity(%onnx::Conv_717)
%onnx::Conv_723 = Identity(%onnx::Conv_717)
%onnx::Conv_720 = Identity(%onnx::Conv_717)
%onnx::Conv_714 = Identity(%onnx::Conv_687)
%onnx::Conv_711 = Identity(%onnx::Conv_687)
%onnx::Conv_708 = Identity(%onnx::Conv_687)
%onnx::Conv_705 = Identity(%onnx::Conv_687)
%onnx::Conv_702 = Identity(%onnx::Conv_687)
%onnx::Conv_699 = Identity(%onnx::Conv_687)
%onnx::Conv_696 = Identity(%onnx::Conv_687)
%onnx::Conv_693 = Identity(%onnx::Conv_636)
%onnx::Conv_690 = Identity(%onnx::Conv_636)
%onnx::Conv_684 = Identity(%onnx::Conv_651)
%onnx::Conv_681 = Identity(%onnx::Conv_651)
%onnx::Conv_678 = Identity(%onnx::Conv_651)
%onnx::Conv_675 = Identity(%onnx::Conv_654)
%onnx::Conv_672 = Identity(%onnx::Conv_654)
%onnx::Conv_669 = Identity(%onnx::Conv_651)
%onnx::Conv_666 = Identity(%onnx::Conv_651)
%onnx::Conv_663 = Identity(%onnx::Conv_651)
%onnx::Conv_660 = Identity(%onnx::Conv_651)
%onnx::Conv_657 = Identity(%onnx::Conv_654)
%onnx::Conv_648 = Identity(%onnx::Conv_633)
%onnx::Conv_645 = Identity(%onnx::Conv_633)
%onnx::Conv_642 = Identity(%onnx::Conv_633)
%onnx::Conv_639 = Identity(%onnx::Conv_636)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_632, %onnx::Conv_633)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.8/Add_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.9/relu/Relu_output_0 = Relu(%/cells.9/conv/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/relu/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.9/relu/Relu_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_803, %onnx::Conv_804)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_806, %onnx::Conv_807)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_809, %onnx::Conv_810)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_812, %onnx::Conv_813)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_815, %onnx::Conv_816)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_818, %onnx::Conv_819)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%630 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %630
}
|
val_accuracy
| 0
| 69,028,736
| 1,839,092
|
{'zcp_synflow': 80.19092579604254, 'zcp_zen': 68.9868392944336, 'zcp_epe_nas': 6.339754348755752, 'zcp_fisher': 0.16472576558589935, 'zcp_flops': 69028736.0, 'zcp_grad_norm': 27.421968460083008, 'zcp_grasp': -0.2770195007324219, 'zcp_jacov': -16.07264913338746, 'zcp_l2_norm': 631.8302001953125, 'zcp_nwot': 211.47568725794602, 'zcp_params': 1839092.0, 'zcp_plain': -0.00015873112715780735, 'zcp_snip': 51.55971145629883, 'lat_1080ti_1': 0.5880878418285451, 'lat_1080ti_32': 0.5868427370626829, 'lat_1080ti_64': 0.47267222400587733, 'lat_2080ti_1': 0.6017082106158884, 'lat_2080ti_32': 0.5548781527132874, 'lat_2080ti_64': 0.4433394167714932, 'lat_essential_ph_1': 0.6037735849056604, 'lat_eyeriss': 0.4264023618726276, 'lat_fpga': 0.4843705162330915, 'lat_gold_6226': 0.3176668938225572, 'lat_gold_6240': 0.46306555071816985, 'lat_pixel2': 0.41304347826086957, 'lat_pixel3': 0.4531473384950117, 'lat_raspi4': 0.4870226016081837, 'lat_samsung_a50': 0.16842105263157894, 'lat_samsung_s7': 0.2047244094488189, 'lat_silver_4114': 0.499837519262036, 'lat_silver_4210r': 0.4883711166222054, 'lat_titan_rtx_1': 0.5771831582873415, 'lat_titan_rtx_32': 0.5537021977900904, 'lat_titan_rtx_64': 0.48265472874597903, 'lat_titanx_1': 0.3057518048238198, 'lat_titanx_32': 0.5098616083061156, 'lat_titanx_64': 0.4361952966284695, 'lat_titanxp_1': 0.5416987820957976, 'lat_titanxp_32': 0.5376047394358081, 'lat_titanxp_64': 0.46870043655237076}
| |
FBNet_4374
|
FBNet
|
4374
|
4374
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_579[FLOAT, 16x3x3x3]
%onnx::Conv_580[FLOAT, 16]
%onnx::Conv_582[FLOAT, 48x16x1x1]
%onnx::Conv_583[FLOAT, 48]
%onnx::Conv_585[FLOAT, 48x1x5x5]
%onnx::Conv_588[FLOAT, 16x48x1x1]
%onnx::Conv_591[FLOAT, 24x16x1x1]
%onnx::Conv_592[FLOAT, 24]
%onnx::Conv_594[FLOAT, 24x12x1x1]
%onnx::Conv_597[FLOAT, 24x1x3x3]
%onnx::Conv_600[FLOAT, 24x12x1x1]
%onnx::Conv_603[FLOAT, 144x24x1x1]
%onnx::Conv_604[FLOAT, 144]
%onnx::Conv_606[FLOAT, 144x1x5x5]
%onnx::Conv_609[FLOAT, 24x144x1x1]
%onnx::Conv_612[FLOAT, 72x24x1x1]
%onnx::Conv_613[FLOAT, 72]
%onnx::Conv_615[FLOAT, 72x1x3x3]
%onnx::Conv_618[FLOAT, 32x72x1x1]
%onnx::Conv_619[FLOAT, 32]
%onnx::Conv_621[FLOAT, 192x32x1x1]
%onnx::Conv_622[FLOAT, 192]
%onnx::Conv_624[FLOAT, 192x1x5x5]
%onnx::Conv_627[FLOAT, 32x192x1x1]
%onnx::Conv_630[FLOAT, 32x32x1x1]
%onnx::Conv_633[FLOAT, 32x1x3x3]
%onnx::Conv_636[FLOAT, 32x32x1x1]
%onnx::Conv_639[FLOAT, 192x32x1x1]
%onnx::Conv_642[FLOAT, 192x1x3x3]
%onnx::Conv_645[FLOAT, 32x192x1x1]
%onnx::Conv_648[FLOAT, 64x32x1x1]
%onnx::Conv_649[FLOAT, 64]
%onnx::Conv_651[FLOAT, 64x64x1x1]
%onnx::Conv_654[FLOAT, 64x1x5x5]
%onnx::Conv_657[FLOAT, 64x64x1x1]
%onnx::Conv_660[FLOAT, 64x32x1x1]
%onnx::Conv_663[FLOAT, 64x1x5x5]
%onnx::Conv_666[FLOAT, 64x32x1x1]
%onnx::Conv_669[FLOAT, 384x64x1x1]
%onnx::Conv_670[FLOAT, 384]
%onnx::Conv_672[FLOAT, 384x1x3x3]
%onnx::Conv_675[FLOAT, 64x384x1x1]
%onnx::Conv_678[FLOAT, 64x32x1x1]
%onnx::Conv_681[FLOAT, 64x1x3x3]
%onnx::Conv_684[FLOAT, 112x32x1x1]
%onnx::Conv_685[FLOAT, 112]
%onnx::Conv_687[FLOAT, 336x112x1x1]
%onnx::Conv_688[FLOAT, 336]
%onnx::Conv_690[FLOAT, 336x1x5x5]
%onnx::Conv_693[FLOAT, 112x336x1x1]
%onnx::Conv_696[FLOAT, 336x112x1x1]
%onnx::Conv_699[FLOAT, 336x1x3x3]
%onnx::Conv_702[FLOAT, 112x336x1x1]
%onnx::Conv_705[FLOAT, 112x56x1x1]
%onnx::Conv_708[FLOAT, 112x1x3x3]
%onnx::Conv_711[FLOAT, 112x56x1x1]
%onnx::Conv_714[FLOAT, 672x112x1x1]
%onnx::Conv_715[FLOAT, 672]
%onnx::Conv_717[FLOAT, 672x1x3x3]
%onnx::Conv_720[FLOAT, 184x672x1x1]
%onnx::Conv_721[FLOAT, 184]
%onnx::Conv_723[FLOAT, 552x184x1x1]
%onnx::Conv_724[FLOAT, 552]
%onnx::Conv_726[FLOAT, 552x1x3x3]
%onnx::Conv_729[FLOAT, 184x552x1x1]
%onnx::Conv_732[FLOAT, 1104x184x1x1]
%onnx::Conv_733[FLOAT, 1104]
%onnx::Conv_735[FLOAT, 1104x1x3x3]
%onnx::Conv_738[FLOAT, 352x1104x1x1]
%onnx::Conv_739[FLOAT, 352]
%onnx::Conv_741[FLOAT, 1504x352x1x1]
%onnx::Conv_742[FLOAT, 1504]
) {
%onnx::Conv_736 = Identity(%onnx::Conv_733)
%onnx::Conv_730 = Identity(%onnx::Conv_721)
%onnx::Conv_727 = Identity(%onnx::Conv_724)
%onnx::Conv_718 = Identity(%onnx::Conv_715)
%onnx::Conv_712 = Identity(%onnx::Conv_685)
%onnx::Conv_709 = Identity(%onnx::Conv_685)
%onnx::Conv_706 = Identity(%onnx::Conv_685)
%onnx::Conv_703 = Identity(%onnx::Conv_685)
%onnx::Conv_700 = Identity(%onnx::Conv_688)
%onnx::Conv_697 = Identity(%onnx::Conv_688)
%onnx::Conv_694 = Identity(%onnx::Conv_685)
%onnx::Conv_691 = Identity(%onnx::Conv_688)
%onnx::Conv_682 = Identity(%onnx::Conv_649)
%onnx::Conv_679 = Identity(%onnx::Conv_649)
%onnx::Conv_676 = Identity(%onnx::Conv_649)
%onnx::Conv_673 = Identity(%onnx::Conv_670)
%onnx::Conv_667 = Identity(%onnx::Conv_649)
%onnx::Conv_664 = Identity(%onnx::Conv_649)
%onnx::Conv_661 = Identity(%onnx::Conv_649)
%onnx::Conv_658 = Identity(%onnx::Conv_649)
%onnx::Conv_655 = Identity(%onnx::Conv_649)
%onnx::Conv_652 = Identity(%onnx::Conv_649)
%onnx::Conv_646 = Identity(%onnx::Conv_619)
%onnx::Conv_643 = Identity(%onnx::Conv_622)
%onnx::Conv_640 = Identity(%onnx::Conv_622)
%onnx::Conv_637 = Identity(%onnx::Conv_619)
%onnx::Conv_634 = Identity(%onnx::Conv_619)
%onnx::Conv_631 = Identity(%onnx::Conv_619)
%onnx::Conv_628 = Identity(%onnx::Conv_619)
%onnx::Conv_625 = Identity(%onnx::Conv_622)
%onnx::Conv_616 = Identity(%onnx::Conv_613)
%onnx::Conv_610 = Identity(%onnx::Conv_592)
%onnx::Conv_607 = Identity(%onnx::Conv_604)
%onnx::Conv_601 = Identity(%onnx::Conv_592)
%onnx::Conv_598 = Identity(%onnx::Conv_592)
%onnx::Conv_595 = Identity(%onnx::Conv_592)
%onnx::Conv_589 = Identity(%onnx::Conv_580)
%onnx::Conv_586 = Identity(%onnx::Conv_583)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_579, %onnx::Conv_580)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_582, %onnx::Conv_583)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_585, %onnx::Conv_586)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_588, %onnx::Conv_589)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_591, %onnx::Conv_592)
%/cells.1/relu/Relu_output_0 = Relu(%/cells.1/conv/Conv_output_0)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/relu/Relu_output_0, %onnx::Conv_594, %onnx::Conv_595)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_597, %onnx::Conv_598)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_600, %onnx::Conv_601)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/relu/Relu_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_603, %onnx::Conv_604)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_606, %onnx::Conv_607)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_609, %onnx::Conv_610)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_612, %onnx::Conv_613)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_615, %onnx::Conv_616)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_618, %onnx::Conv_619)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_621, %onnx::Conv_622)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_624, %onnx::Conv_625)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_627, %onnx::Conv_628)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_630, %onnx::Conv_631)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.8/Add_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.9/relu/Relu_output_0 = Relu(%/cells.9/conv/Conv_output_0)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/relu/Relu_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/relu/Relu_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%577 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %577
}
|
val_accuracy
| 0
| 73,946,752
| 2,023,372
|
{'zcp_synflow': 69.03154363737825, 'zcp_zen': 60.49693298339844, 'zcp_epe_nas': 20.057679059601966, 'zcp_fisher': 0.07751917839050293, 'zcp_flops': 73946752.0, 'zcp_grad_norm': 20.557844161987305, 'zcp_grasp': -0.2319316864013672, 'zcp_jacov': -16.05554273393244, 'zcp_l2_norm': 571.9319458007812, 'zcp_nwot': 213.21009923205276, 'zcp_params': 2023372.0, 'zcp_plain': 0.0070227449759840965, 'zcp_snip': 38.918704986572266, 'lat_1080ti_1': 0.39800991163161864, 'lat_1080ti_32': 0.42489371571566076, 'lat_1080ti_64': 0.38061255769594493, 'lat_2080ti_1': 0.3899507302254854, 'lat_2080ti_32': 0.3876475478840673, 'lat_2080ti_64': 0.411208548062937, 'lat_essential_ph_1': 0.32075471698113206, 'lat_eyeriss': 0.477655731392001, 'lat_fpga': 0.4922569983870808, 'lat_gold_6226': 0.384472982722824, 'lat_gold_6240': 0.3906416034815428, 'lat_pixel2': 0.41304347826086957, 'lat_pixel3': 0.5014275336532211, 'lat_raspi4': 0.5240094948443538, 'lat_samsung_a50': 0.21052631578947367, 'lat_samsung_s7': 0.11811023622047244, 'lat_silver_4114': 0.42428613712758484, 'lat_silver_4210r': 0.39549463297414705, 'lat_titan_rtx_1': 0.35940128765224905, 'lat_titan_rtx_32': 0.3621563539377202, 'lat_titan_rtx_64': 0.3832401811038131, 'lat_titanx_1': 0.1929854681554032, 'lat_titanx_32': 0.36872140603951986, 'lat_titanx_64': 0.39820772574891383, 'lat_titanxp_1': 0.3500606261681864, 'lat_titanxp_32': 0.3664698632965715, 'lat_titanxp_64': 0.3827626491700871}
| |
FBNet_3080
|
FBNet
|
3080
|
3080
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_750[FLOAT, 16x3x3x3]
%onnx::Conv_751[FLOAT, 16]
%onnx::Conv_753[FLOAT, 16x8x1x1]
%onnx::Conv_756[FLOAT, 16x1x3x3]
%onnx::Conv_759[FLOAT, 24x8x1x1]
%onnx::Conv_760[FLOAT, 24]
%onnx::Conv_762[FLOAT, 24x24x1x1]
%onnx::Conv_765[FLOAT, 24x1x3x3]
%onnx::Conv_768[FLOAT, 24x24x1x1]
%onnx::Conv_771[FLOAT, 144x24x1x1]
%onnx::Conv_772[FLOAT, 144]
%onnx::Conv_774[FLOAT, 144x1x3x3]
%onnx::Conv_777[FLOAT, 24x144x1x1]
%onnx::Conv_780[FLOAT, 24x12x1x1]
%onnx::Conv_783[FLOAT, 24x1x5x5]
%onnx::Conv_786[FLOAT, 24x12x1x1]
%onnx::Conv_789[FLOAT, 24x12x1x1]
%onnx::Conv_792[FLOAT, 24x1x3x3]
%onnx::Conv_795[FLOAT, 32x12x1x1]
%onnx::Conv_796[FLOAT, 32]
%onnx::Conv_798[FLOAT, 32x16x1x1]
%onnx::Conv_801[FLOAT, 32x1x3x3]
%onnx::Conv_804[FLOAT, 32x16x1x1]
%onnx::Conv_807[FLOAT, 32x16x1x1]
%onnx::Conv_810[FLOAT, 32x1x3x3]
%onnx::Conv_813[FLOAT, 32x16x1x1]
%onnx::Conv_816[FLOAT, 32x16x1x1]
%onnx::Conv_819[FLOAT, 32x1x3x3]
%onnx::Conv_822[FLOAT, 32x16x1x1]
%onnx::Conv_825[FLOAT, 32x32x1x1]
%onnx::Conv_828[FLOAT, 32x1x3x3]
%onnx::Conv_831[FLOAT, 64x32x1x1]
%onnx::Conv_832[FLOAT, 64]
%onnx::Conv_834[FLOAT, 192x64x1x1]
%onnx::Conv_835[FLOAT, 192]
%onnx::Conv_837[FLOAT, 192x1x3x3]
%onnx::Conv_840[FLOAT, 64x192x1x1]
%onnx::Conv_843[FLOAT, 64x64x1x1]
%onnx::Conv_846[FLOAT, 64x1x5x5]
%onnx::Conv_849[FLOAT, 64x64x1x1]
%onnx::Conv_852[FLOAT, 64x32x1x1]
%onnx::Conv_855[FLOAT, 64x1x5x5]
%onnx::Conv_858[FLOAT, 64x32x1x1]
%onnx::Conv_861[FLOAT, 64x32x1x1]
%onnx::Conv_864[FLOAT, 64x1x5x5]
%onnx::Conv_867[FLOAT, 112x32x1x1]
%onnx::Conv_868[FLOAT, 112]
%onnx::Conv_870[FLOAT, 112x112x1x1]
%onnx::Conv_873[FLOAT, 112x1x5x5]
%onnx::Conv_876[FLOAT, 112x112x1x1]
%onnx::Conv_879[FLOAT, 672x112x1x1]
%onnx::Conv_880[FLOAT, 672]
%onnx::Conv_882[FLOAT, 672x1x3x3]
%onnx::Conv_885[FLOAT, 112x672x1x1]
%onnx::Conv_888[FLOAT, 672x112x1x1]
%onnx::Conv_891[FLOAT, 672x1x5x5]
%onnx::Conv_894[FLOAT, 112x672x1x1]
%onnx::Conv_897[FLOAT, 112x112x1x1]
%onnx::Conv_900[FLOAT, 112x1x5x5]
%onnx::Conv_903[FLOAT, 184x112x1x1]
%onnx::Conv_904[FLOAT, 184]
%onnx::Conv_906[FLOAT, 184x184x1x1]
%onnx::Conv_909[FLOAT, 184x1x5x5]
%onnx::Conv_912[FLOAT, 184x184x1x1]
%onnx::Conv_915[FLOAT, 184x92x1x1]
%onnx::Conv_918[FLOAT, 184x1x5x5]
%onnx::Conv_921[FLOAT, 184x92x1x1]
%onnx::Conv_924[FLOAT, 184x92x1x1]
%onnx::Conv_927[FLOAT, 184x1x5x5]
%onnx::Conv_930[FLOAT, 352x92x1x1]
%onnx::Conv_931[FLOAT, 352]
%onnx::Conv_933[FLOAT, 1504x352x1x1]
%onnx::Conv_934[FLOAT, 1504]
) {
%onnx::Conv_928 = Identity(%onnx::Conv_904)
%onnx::Conv_925 = Identity(%onnx::Conv_904)
%onnx::Conv_922 = Identity(%onnx::Conv_904)
%onnx::Conv_919 = Identity(%onnx::Conv_904)
%onnx::Conv_916 = Identity(%onnx::Conv_904)
%onnx::Conv_913 = Identity(%onnx::Conv_904)
%onnx::Conv_910 = Identity(%onnx::Conv_904)
%onnx::Conv_907 = Identity(%onnx::Conv_904)
%onnx::Conv_901 = Identity(%onnx::Conv_868)
%onnx::Conv_898 = Identity(%onnx::Conv_868)
%onnx::Conv_895 = Identity(%onnx::Conv_868)
%onnx::Conv_892 = Identity(%onnx::Conv_880)
%onnx::Conv_889 = Identity(%onnx::Conv_880)
%onnx::Conv_886 = Identity(%onnx::Conv_868)
%onnx::Conv_883 = Identity(%onnx::Conv_880)
%onnx::Conv_877 = Identity(%onnx::Conv_868)
%onnx::Conv_874 = Identity(%onnx::Conv_868)
%onnx::Conv_871 = Identity(%onnx::Conv_868)
%onnx::Conv_865 = Identity(%onnx::Conv_832)
%onnx::Conv_862 = Identity(%onnx::Conv_832)
%onnx::Conv_859 = Identity(%onnx::Conv_832)
%onnx::Conv_856 = Identity(%onnx::Conv_832)
%onnx::Conv_853 = Identity(%onnx::Conv_832)
%onnx::Conv_850 = Identity(%onnx::Conv_832)
%onnx::Conv_847 = Identity(%onnx::Conv_832)
%onnx::Conv_844 = Identity(%onnx::Conv_832)
%onnx::Conv_841 = Identity(%onnx::Conv_832)
%onnx::Conv_838 = Identity(%onnx::Conv_835)
%onnx::Conv_829 = Identity(%onnx::Conv_796)
%onnx::Conv_826 = Identity(%onnx::Conv_796)
%onnx::Conv_823 = Identity(%onnx::Conv_796)
%onnx::Conv_820 = Identity(%onnx::Conv_796)
%onnx::Conv_817 = Identity(%onnx::Conv_796)
%onnx::Conv_814 = Identity(%onnx::Conv_796)
%onnx::Conv_811 = Identity(%onnx::Conv_796)
%onnx::Conv_808 = Identity(%onnx::Conv_796)
%onnx::Conv_805 = Identity(%onnx::Conv_796)
%onnx::Conv_802 = Identity(%onnx::Conv_796)
%onnx::Conv_799 = Identity(%onnx::Conv_796)
%onnx::Conv_793 = Identity(%onnx::Conv_760)
%onnx::Conv_790 = Identity(%onnx::Conv_760)
%onnx::Conv_787 = Identity(%onnx::Conv_760)
%onnx::Conv_784 = Identity(%onnx::Conv_760)
%onnx::Conv_781 = Identity(%onnx::Conv_760)
%onnx::Conv_778 = Identity(%onnx::Conv_760)
%onnx::Conv_775 = Identity(%onnx::Conv_772)
%onnx::Conv_769 = Identity(%onnx::Conv_760)
%onnx::Conv_766 = Identity(%onnx::Conv_760)
%onnx::Conv_763 = Identity(%onnx::Conv_760)
%onnx::Conv_757 = Identity(%onnx::Conv_751)
%onnx::Conv_754 = Identity(%onnx::Conv_751)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_750, %onnx::Conv_751)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_858, %onnx::Conv_859)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_861, %onnx::Conv_862)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_864, %onnx::Conv_865)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_867, %onnx::Conv_868)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_870, %onnx::Conv_871)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_873, %onnx::Conv_874)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_876, %onnx::Conv_877)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_879, %onnx::Conv_880)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_882, %onnx::Conv_883)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_885, %onnx::Conv_886)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_888, %onnx::Conv_889)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_891, %onnx::Conv_892)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_894, %onnx::Conv_895)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_897, %onnx::Conv_898)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_900, %onnx::Conv_901)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_903, %onnx::Conv_904)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_906, %onnx::Conv_907)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_909, %onnx::Conv_910)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_912, %onnx::Conv_913)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_915, %onnx::Conv_916)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_918, %onnx::Conv_919)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_921, %onnx::Conv_922)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_924, %onnx::Conv_925)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_927, %onnx::Conv_928)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_930, %onnx::Conv_931)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_933, %onnx::Conv_934)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%748 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %748
}
|
val_accuracy
| 0
| 54,337,920
| 1,319,348
|
{'zcp_synflow': 68.45325126346941, 'zcp_zen': 60.62423324584961, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.09284783899784088, 'zcp_flops': 54337920.0, 'zcp_grad_norm': 19.841983795166016, 'zcp_grasp': -0.0019178390502929688, 'zcp_jacov': -16.048937975127174, 'zcp_l2_norm': 513.8184204101562, 'zcp_nwot': 208.29284829615654, 'zcp_params': 1319348.0, 'zcp_plain': -0.00022451314725913107, 'zcp_snip': 29.099151611328125, 'lat_1080ti_1': 0.7453994067246493, 'lat_1080ti_32': 0.6393370921206525, 'lat_1080ti_64': 0.41613411310051135, 'lat_2080ti_1': 0.7438325241893997, 'lat_2080ti_32': 0.655181512484331, 'lat_2080ti_64': 0.465702718295325, 'lat_essential_ph_1': 0.20754716981132076, 'lat_eyeriss': 0.21404470687473653, 'lat_fpga': 0.31625090293788777, 'lat_gold_6226': 0.16311841483534076, 'lat_gold_6240': 0.3737368553144262, 'lat_pixel2': 0.17391304347826086, 'lat_pixel3': 0.27626738420435243, 'lat_raspi4': 0.2892726492903232, 'lat_samsung_a50': 0.10526315789473684, 'lat_samsung_s7': 0.06299212598425197, 'lat_silver_4114': 0.4292527175586989, 'lat_silver_4210r': 0.46984553322539624, 'lat_titan_rtx_1': 0.6859404360932928, 'lat_titan_rtx_32': 0.6278875819411106, 'lat_titan_rtx_64': 0.5023810969020109, 'lat_titanx_1': 0.35737520258234884, 'lat_titanx_32': 0.5247790813754623, 'lat_titanx_64': 0.38315674417455514, 'lat_titanxp_1': 0.6544847229209741, 'lat_titanxp_32': 0.5741836133981744, 'lat_titanxp_64': 0.4404519391135153}
| |
FBNet_6
|
FBNet
|
6
|
6
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_630[FLOAT, 16x3x3x3]
%onnx::Conv_631[FLOAT, 16]
%onnx::Conv_633[FLOAT, 16x8x1x1]
%onnx::Conv_636[FLOAT, 16x1x5x5]
%onnx::Conv_639[FLOAT, 16x8x1x1]
%onnx::Conv_642[FLOAT, 48x16x1x1]
%onnx::Conv_643[FLOAT, 48]
%onnx::Conv_645[FLOAT, 48x1x3x3]
%onnx::Conv_648[FLOAT, 24x48x1x1]
%onnx::Conv_649[FLOAT, 24]
%onnx::Conv_651[FLOAT, 144x24x1x1]
%onnx::Conv_652[FLOAT, 144]
%onnx::Conv_654[FLOAT, 144x1x5x5]
%onnx::Conv_657[FLOAT, 24x144x1x1]
%onnx::Conv_660[FLOAT, 72x24x1x1]
%onnx::Conv_661[FLOAT, 72]
%onnx::Conv_663[FLOAT, 72x1x3x3]
%onnx::Conv_666[FLOAT, 24x72x1x1]
%onnx::Conv_669[FLOAT, 24x12x1x1]
%onnx::Conv_672[FLOAT, 24x1x5x5]
%onnx::Conv_675[FLOAT, 24x12x1x1]
%onnx::Conv_678[FLOAT, 24x24x1x1]
%onnx::Conv_681[FLOAT, 24x1x3x3]
%onnx::Conv_684[FLOAT, 32x24x1x1]
%onnx::Conv_685[FLOAT, 32]
%onnx::Conv_687[FLOAT, 32x32x1x1]
%onnx::Conv_690[FLOAT, 32x1x3x3]
%onnx::Conv_693[FLOAT, 32x32x1x1]
%onnx::Conv_696[FLOAT, 192x32x1x1]
%onnx::Conv_697[FLOAT, 192]
%onnx::Conv_699[FLOAT, 192x1x3x3]
%onnx::Conv_702[FLOAT, 32x192x1x1]
%onnx::Conv_705[FLOAT, 96x32x1x1]
%onnx::Conv_706[FLOAT, 96]
%onnx::Conv_708[FLOAT, 96x1x5x5]
%onnx::Conv_711[FLOAT, 64x96x1x1]
%onnx::Conv_712[FLOAT, 64]
%onnx::Conv_714[FLOAT, 64x64x1x1]
%onnx::Conv_717[FLOAT, 64x1x5x5]
%onnx::Conv_720[FLOAT, 64x64x1x1]
%onnx::Conv_723[FLOAT, 384x64x1x1]
%onnx::Conv_724[FLOAT, 384]
%onnx::Conv_726[FLOAT, 384x1x3x3]
%onnx::Conv_729[FLOAT, 64x384x1x1]
%onnx::Conv_732[FLOAT, 64x64x1x1]
%onnx::Conv_735[FLOAT, 64x1x5x5]
%onnx::Conv_738[FLOAT, 64x64x1x1]
%onnx::Conv_741[FLOAT, 192x64x1x1]
%onnx::Conv_744[FLOAT, 192x1x5x5]
%onnx::Conv_747[FLOAT, 112x192x1x1]
%onnx::Conv_748[FLOAT, 112]
%onnx::Conv_750[FLOAT, 672x112x1x1]
%onnx::Conv_751[FLOAT, 672]
%onnx::Conv_753[FLOAT, 672x1x5x5]
%onnx::Conv_756[FLOAT, 112x672x1x1]
%onnx::Conv_759[FLOAT, 336x112x1x1]
%onnx::Conv_760[FLOAT, 336]
%onnx::Conv_762[FLOAT, 336x1x5x5]
%onnx::Conv_765[FLOAT, 112x336x1x1]
%onnx::Conv_768[FLOAT, 112x112x1x1]
%onnx::Conv_771[FLOAT, 112x1x3x3]
%onnx::Conv_774[FLOAT, 112x112x1x1]
%onnx::Conv_777[FLOAT, 112x112x1x1]
%onnx::Conv_780[FLOAT, 112x1x5x5]
%onnx::Conv_783[FLOAT, 184x112x1x1]
%onnx::Conv_784[FLOAT, 184]
%onnx::Conv_786[FLOAT, 552x184x1x1]
%onnx::Conv_787[FLOAT, 552]
%onnx::Conv_789[FLOAT, 552x1x5x5]
%onnx::Conv_792[FLOAT, 184x552x1x1]
%onnx::Conv_795[FLOAT, 552x184x1x1]
%onnx::Conv_798[FLOAT, 552x1x3x3]
%onnx::Conv_801[FLOAT, 184x552x1x1]
%onnx::Conv_804[FLOAT, 184x184x1x1]
%onnx::Conv_807[FLOAT, 184x1x3x3]
%onnx::Conv_810[FLOAT, 184x184x1x1]
%onnx::Conv_813[FLOAT, 184x184x1x1]
%onnx::Conv_816[FLOAT, 184x1x3x3]
%onnx::Conv_819[FLOAT, 352x184x1x1]
%onnx::Conv_820[FLOAT, 352]
%onnx::Conv_822[FLOAT, 1504x352x1x1]
%onnx::Conv_823[FLOAT, 1504]
) {
%onnx::Conv_817 = Identity(%onnx::Conv_784)
%onnx::Conv_814 = Identity(%onnx::Conv_784)
%onnx::Conv_811 = Identity(%onnx::Conv_784)
%onnx::Conv_808 = Identity(%onnx::Conv_784)
%onnx::Conv_805 = Identity(%onnx::Conv_784)
%onnx::Conv_802 = Identity(%onnx::Conv_784)
%onnx::Conv_799 = Identity(%onnx::Conv_787)
%onnx::Conv_796 = Identity(%onnx::Conv_787)
%onnx::Conv_793 = Identity(%onnx::Conv_784)
%onnx::Conv_790 = Identity(%onnx::Conv_787)
%onnx::Conv_781 = Identity(%onnx::Conv_748)
%onnx::Conv_778 = Identity(%onnx::Conv_748)
%onnx::Conv_775 = Identity(%onnx::Conv_748)
%onnx::Conv_772 = Identity(%onnx::Conv_748)
%onnx::Conv_769 = Identity(%onnx::Conv_748)
%onnx::Conv_766 = Identity(%onnx::Conv_748)
%onnx::Conv_763 = Identity(%onnx::Conv_760)
%onnx::Conv_757 = Identity(%onnx::Conv_748)
%onnx::Conv_754 = Identity(%onnx::Conv_751)
%onnx::Conv_745 = Identity(%onnx::Conv_697)
%onnx::Conv_742 = Identity(%onnx::Conv_697)
%onnx::Conv_739 = Identity(%onnx::Conv_712)
%onnx::Conv_736 = Identity(%onnx::Conv_712)
%onnx::Conv_733 = Identity(%onnx::Conv_712)
%onnx::Conv_730 = Identity(%onnx::Conv_712)
%onnx::Conv_727 = Identity(%onnx::Conv_724)
%onnx::Conv_721 = Identity(%onnx::Conv_712)
%onnx::Conv_718 = Identity(%onnx::Conv_712)
%onnx::Conv_715 = Identity(%onnx::Conv_712)
%onnx::Conv_709 = Identity(%onnx::Conv_706)
%onnx::Conv_703 = Identity(%onnx::Conv_685)
%onnx::Conv_700 = Identity(%onnx::Conv_697)
%onnx::Conv_694 = Identity(%onnx::Conv_685)
%onnx::Conv_691 = Identity(%onnx::Conv_685)
%onnx::Conv_688 = Identity(%onnx::Conv_685)
%onnx::Conv_682 = Identity(%onnx::Conv_649)
%onnx::Conv_679 = Identity(%onnx::Conv_649)
%onnx::Conv_676 = Identity(%onnx::Conv_649)
%onnx::Conv_673 = Identity(%onnx::Conv_649)
%onnx::Conv_670 = Identity(%onnx::Conv_649)
%onnx::Conv_667 = Identity(%onnx::Conv_649)
%onnx::Conv_664 = Identity(%onnx::Conv_661)
%onnx::Conv_658 = Identity(%onnx::Conv_649)
%onnx::Conv_655 = Identity(%onnx::Conv_652)
%onnx::Conv_646 = Identity(%onnx::Conv_643)
%onnx::Conv_640 = Identity(%onnx::Conv_631)
%onnx::Conv_637 = Identity(%onnx::Conv_631)
%onnx::Conv_634 = Identity(%onnx::Conv_631)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_630, %onnx::Conv_631)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%628 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %628
}
|
val_accuracy
| 0
| 73,440,128
| 1,770,444
|
{'zcp_synflow': 84.0171595346383, 'zcp_zen': 72.71915435791016, 'zcp_epe_nas': 18.822925589729927, 'zcp_fisher': 0.13278678059577942, 'zcp_flops': 73440128.0, 'zcp_grad_norm': 25.842756271362305, 'zcp_grasp': 0.07021713256835938, 'zcp_jacov': -16.06000910930099, 'zcp_l2_norm': 663.7381591796875, 'zcp_nwot': 213.56994574978302, 'zcp_params': 1770444.0, 'zcp_plain': 0.0026476748753339052, 'zcp_snip': 46.73044967651367, 'lat_1080ti_1': 0.6006282092250005, 'lat_1080ti_32': 0.5461453190638064, 'lat_1080ti_64': 0.503163873163419, 'lat_2080ti_1': 0.7334710247968601, 'lat_2080ti_32': 0.582002653438664, 'lat_2080ti_64': 0.52384770536578, 'lat_essential_ph_1': 0.2641509433962264, 'lat_eyeriss': 0.5003209066070087, 'lat_fpga': 0.5398826428126142, 'lat_gold_6226': 0.46323404451389544, 'lat_gold_6240': 0.6196295849936064, 'lat_pixel2': 0.2826086956521739, 'lat_pixel3': 0.5170128122167442, 'lat_raspi4': 0.47963014385193664, 'lat_samsung_a50': 0.7368421052631579, 'lat_samsung_s7': 0.6377952755905512, 'lat_silver_4114': 0.5689334402455053, 'lat_silver_4210r': 0.5589819939426226, 'lat_titan_rtx_1': 0.6076896657111691, 'lat_titan_rtx_32': 0.5986837715579153, 'lat_titan_rtx_64': 0.602692658861029, 'lat_titanx_1': 0.3277246340202722, 'lat_titanx_32': 0.5391300881777619, 'lat_titanx_64': 0.4889311526481714, 'lat_titanxp_1': 0.5805558995996509, 'lat_titanxp_32': 0.5940920293583879, 'lat_titanxp_64': 0.5566079907939114}
| |
FBNet_1055
|
FBNet
|
1055
|
1055
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_613[FLOAT, 16x3x3x3]
%onnx::Conv_614[FLOAT, 16]
%onnx::Conv_616[FLOAT, 16x16x1x1]
%onnx::Conv_619[FLOAT, 16x1x5x5]
%onnx::Conv_622[FLOAT, 16x16x1x1]
%onnx::Conv_625[FLOAT, 16x8x1x1]
%onnx::Conv_628[FLOAT, 16x1x3x3]
%onnx::Conv_631[FLOAT, 24x8x1x1]
%onnx::Conv_632[FLOAT, 24]
%onnx::Conv_634[FLOAT, 72x24x1x1]
%onnx::Conv_635[FLOAT, 72]
%onnx::Conv_637[FLOAT, 72x1x3x3]
%onnx::Conv_640[FLOAT, 24x72x1x1]
%onnx::Conv_643[FLOAT, 72x24x1x1]
%onnx::Conv_646[FLOAT, 72x1x3x3]
%onnx::Conv_649[FLOAT, 24x72x1x1]
%onnx::Conv_652[FLOAT, 24x24x1x1]
%onnx::Conv_655[FLOAT, 24x1x5x5]
%onnx::Conv_658[FLOAT, 32x24x1x1]
%onnx::Conv_659[FLOAT, 32]
%onnx::Conv_661[FLOAT, 32x16x1x1]
%onnx::Conv_664[FLOAT, 32x1x3x3]
%onnx::Conv_667[FLOAT, 32x16x1x1]
%onnx::Conv_670[FLOAT, 32x32x1x1]
%onnx::Conv_673[FLOAT, 32x1x5x5]
%onnx::Conv_676[FLOAT, 32x32x1x1]
%onnx::Conv_679[FLOAT, 96x32x1x1]
%onnx::Conv_680[FLOAT, 96]
%onnx::Conv_682[FLOAT, 96x1x3x3]
%onnx::Conv_685[FLOAT, 64x96x1x1]
%onnx::Conv_686[FLOAT, 64]
%onnx::Conv_688[FLOAT, 384x64x1x1]
%onnx::Conv_689[FLOAT, 384]
%onnx::Conv_691[FLOAT, 384x1x3x3]
%onnx::Conv_694[FLOAT, 64x384x1x1]
%onnx::Conv_697[FLOAT, 64x64x1x1]
%onnx::Conv_700[FLOAT, 64x1x3x3]
%onnx::Conv_703[FLOAT, 64x64x1x1]
%onnx::Conv_706[FLOAT, 192x64x1x1]
%onnx::Conv_707[FLOAT, 192]
%onnx::Conv_709[FLOAT, 192x1x3x3]
%onnx::Conv_712[FLOAT, 112x192x1x1]
%onnx::Conv_713[FLOAT, 112]
%onnx::Conv_715[FLOAT, 112x112x1x1]
%onnx::Conv_718[FLOAT, 112x1x5x5]
%onnx::Conv_721[FLOAT, 112x112x1x1]
%onnx::Conv_724[FLOAT, 672x112x1x1]
%onnx::Conv_725[FLOAT, 672]
%onnx::Conv_727[FLOAT, 672x1x5x5]
%onnx::Conv_730[FLOAT, 112x672x1x1]
%onnx::Conv_733[FLOAT, 336x112x1x1]
%onnx::Conv_734[FLOAT, 336]
%onnx::Conv_736[FLOAT, 336x1x5x5]
%onnx::Conv_739[FLOAT, 112x336x1x1]
%onnx::Conv_742[FLOAT, 672x112x1x1]
%onnx::Conv_745[FLOAT, 672x1x3x3]
%onnx::Conv_748[FLOAT, 184x672x1x1]
%onnx::Conv_749[FLOAT, 184]
%onnx::Conv_751[FLOAT, 184x184x1x1]
%onnx::Conv_754[FLOAT, 184x1x5x5]
%onnx::Conv_757[FLOAT, 184x184x1x1]
%onnx::Conv_760[FLOAT, 184x184x1x1]
%onnx::Conv_763[FLOAT, 184x1x3x3]
%onnx::Conv_766[FLOAT, 184x184x1x1]
%onnx::Conv_769[FLOAT, 184x92x1x1]
%onnx::Conv_772[FLOAT, 184x1x5x5]
%onnx::Conv_775[FLOAT, 184x92x1x1]
%onnx::Conv_778[FLOAT, 184x92x1x1]
%onnx::Conv_781[FLOAT, 184x1x3x3]
%onnx::Conv_784[FLOAT, 352x92x1x1]
%onnx::Conv_785[FLOAT, 352]
%onnx::Conv_787[FLOAT, 1504x352x1x1]
%onnx::Conv_788[FLOAT, 1504]
) {
%onnx::Conv_782 = Identity(%onnx::Conv_749)
%onnx::Conv_779 = Identity(%onnx::Conv_749)
%onnx::Conv_776 = Identity(%onnx::Conv_749)
%onnx::Conv_773 = Identity(%onnx::Conv_749)
%onnx::Conv_770 = Identity(%onnx::Conv_749)
%onnx::Conv_767 = Identity(%onnx::Conv_749)
%onnx::Conv_764 = Identity(%onnx::Conv_749)
%onnx::Conv_761 = Identity(%onnx::Conv_749)
%onnx::Conv_758 = Identity(%onnx::Conv_749)
%onnx::Conv_755 = Identity(%onnx::Conv_749)
%onnx::Conv_752 = Identity(%onnx::Conv_749)
%onnx::Conv_746 = Identity(%onnx::Conv_725)
%onnx::Conv_743 = Identity(%onnx::Conv_725)
%onnx::Conv_740 = Identity(%onnx::Conv_713)
%onnx::Conv_737 = Identity(%onnx::Conv_734)
%onnx::Conv_731 = Identity(%onnx::Conv_713)
%onnx::Conv_728 = Identity(%onnx::Conv_725)
%onnx::Conv_722 = Identity(%onnx::Conv_713)
%onnx::Conv_719 = Identity(%onnx::Conv_713)
%onnx::Conv_716 = Identity(%onnx::Conv_713)
%onnx::Conv_710 = Identity(%onnx::Conv_707)
%onnx::Conv_704 = Identity(%onnx::Conv_686)
%onnx::Conv_701 = Identity(%onnx::Conv_686)
%onnx::Conv_698 = Identity(%onnx::Conv_686)
%onnx::Conv_695 = Identity(%onnx::Conv_686)
%onnx::Conv_692 = Identity(%onnx::Conv_689)
%onnx::Conv_683 = Identity(%onnx::Conv_680)
%onnx::Conv_677 = Identity(%onnx::Conv_659)
%onnx::Conv_674 = Identity(%onnx::Conv_659)
%onnx::Conv_671 = Identity(%onnx::Conv_659)
%onnx::Conv_668 = Identity(%onnx::Conv_659)
%onnx::Conv_665 = Identity(%onnx::Conv_659)
%onnx::Conv_662 = Identity(%onnx::Conv_659)
%onnx::Conv_656 = Identity(%onnx::Conv_632)
%onnx::Conv_653 = Identity(%onnx::Conv_632)
%onnx::Conv_650 = Identity(%onnx::Conv_632)
%onnx::Conv_647 = Identity(%onnx::Conv_635)
%onnx::Conv_644 = Identity(%onnx::Conv_635)
%onnx::Conv_641 = Identity(%onnx::Conv_632)
%onnx::Conv_638 = Identity(%onnx::Conv_635)
%onnx::Conv_629 = Identity(%onnx::Conv_614)
%onnx::Conv_626 = Identity(%onnx::Conv_614)
%onnx::Conv_623 = Identity(%onnx::Conv_614)
%onnx::Conv_620 = Identity(%onnx::Conv_614)
%onnx::Conv_617 = Identity(%onnx::Conv_614)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_613, %onnx::Conv_614)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_616, %onnx::Conv_617)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_619, %onnx::Conv_620)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_622, %onnx::Conv_623)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_625, %onnx::Conv_626)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_628, %onnx::Conv_629)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_631, %onnx::Conv_632)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_634, %onnx::Conv_635)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_637, %onnx::Conv_638)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_640, %onnx::Conv_641)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_643, %onnx::Conv_644)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_646, %onnx::Conv_647)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_649, %onnx::Conv_650)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_652, %onnx::Conv_653)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_655, %onnx::Conv_656)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_658, %onnx::Conv_659)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%611 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %611
}
|
val_accuracy
| 0
| 59,106,688
| 1,538,924
|
{'zcp_synflow': 73.09563190957924, 'zcp_zen': 63.636558532714844, 'zcp_epe_nas': 25.93979532588962, 'zcp_fisher': 0.08717767894268036, 'zcp_flops': 59106688.0, 'zcp_grad_norm': 20.773963928222656, 'zcp_grasp': 0.014436721801757812, 'zcp_jacov': -16.066917220129362, 'zcp_l2_norm': 574.3762817382812, 'zcp_nwot': 207.86801414204837, 'zcp_params': 1538924.0, 'zcp_plain': -0.00019139022333547473, 'zcp_snip': 36.42169189453125, 'lat_1080ti_1': 0.4258459261612619, 'lat_1080ti_32': 0.34577868338829276, 'lat_1080ti_64': 0.25985768618723737, 'lat_2080ti_1': 0.48616241381822384, 'lat_2080ti_32': 0.3642863943379229, 'lat_2080ti_64': 0.2819359994645654, 'lat_essential_ph_1': 0.24528301886792453, 'lat_eyeriss': 0.2615205471916088, 'lat_fpga': 0.3297380737984742, 'lat_gold_6226': 0.22352075007526379, 'lat_gold_6240': 0.3233374932619555, 'lat_pixel2': 0.21739130434782608, 'lat_pixel3': 0.2659745279203868, 'lat_raspi4': 0.24799038536656848, 'lat_samsung_a50': 0.11578947368421053, 'lat_samsung_s7': 0.11811023622047244, 'lat_silver_4114': 0.5218522918289371, 'lat_silver_4210r': 0.4056487183491762, 'lat_titan_rtx_1': 0.4558238361275659, 'lat_titan_rtx_32': 0.361222141256893, 'lat_titan_rtx_64': 0.29874152866258347, 'lat_titanx_1': 0.23922056565489797, 'lat_titanx_32': 0.32360139877395183, 'lat_titanx_64': 0.28741245372885293, 'lat_titanxp_1': 0.4206682607939203, 'lat_titanxp_32': 0.3553609655979496, 'lat_titanxp_64': 0.2773742919050708}
| |
FBNet_3616
|
FBNet
|
3616
|
3616
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_669[FLOAT, 16x3x3x3]
%onnx::Conv_670[FLOAT, 16]
%onnx::Conv_672[FLOAT, 96x16x1x1]
%onnx::Conv_673[FLOAT, 96]
%onnx::Conv_675[FLOAT, 96x1x5x5]
%onnx::Conv_678[FLOAT, 16x96x1x1]
%onnx::Conv_681[FLOAT, 48x16x1x1]
%onnx::Conv_682[FLOAT, 48]
%onnx::Conv_684[FLOAT, 48x1x3x3]
%onnx::Conv_687[FLOAT, 24x48x1x1]
%onnx::Conv_688[FLOAT, 24]
%onnx::Conv_690[FLOAT, 24x12x1x1]
%onnx::Conv_693[FLOAT, 24x1x3x3]
%onnx::Conv_696[FLOAT, 24x12x1x1]
%onnx::Conv_699[FLOAT, 144x24x1x1]
%onnx::Conv_700[FLOAT, 144]
%onnx::Conv_702[FLOAT, 144x1x3x3]
%onnx::Conv_705[FLOAT, 24x144x1x1]
%onnx::Conv_708[FLOAT, 144x24x1x1]
%onnx::Conv_711[FLOAT, 144x1x3x3]
%onnx::Conv_714[FLOAT, 24x144x1x1]
%onnx::Conv_717[FLOAT, 144x24x1x1]
%onnx::Conv_720[FLOAT, 144x1x5x5]
%onnx::Conv_723[FLOAT, 32x144x1x1]
%onnx::Conv_724[FLOAT, 32]
%onnx::Conv_726[FLOAT, 32x16x1x1]
%onnx::Conv_729[FLOAT, 32x1x3x3]
%onnx::Conv_732[FLOAT, 32x16x1x1]
%onnx::Conv_735[FLOAT, 96x32x1x1]
%onnx::Conv_738[FLOAT, 96x1x5x5]
%onnx::Conv_741[FLOAT, 32x96x1x1]
%onnx::Conv_744[FLOAT, 96x32x1x1]
%onnx::Conv_747[FLOAT, 96x1x3x3]
%onnx::Conv_750[FLOAT, 32x96x1x1]
%onnx::Conv_753[FLOAT, 32x16x1x1]
%onnx::Conv_756[FLOAT, 32x1x5x5]
%onnx::Conv_759[FLOAT, 64x16x1x1]
%onnx::Conv_760[FLOAT, 64]
%onnx::Conv_762[FLOAT, 384x64x1x1]
%onnx::Conv_763[FLOAT, 384]
%onnx::Conv_765[FLOAT, 384x1x5x5]
%onnx::Conv_768[FLOAT, 64x384x1x1]
%onnx::Conv_771[FLOAT, 64x32x1x1]
%onnx::Conv_774[FLOAT, 64x1x5x5]
%onnx::Conv_777[FLOAT, 64x32x1x1]
%onnx::Conv_780[FLOAT, 192x64x1x1]
%onnx::Conv_781[FLOAT, 192]
%onnx::Conv_783[FLOAT, 192x1x5x5]
%onnx::Conv_786[FLOAT, 112x192x1x1]
%onnx::Conv_787[FLOAT, 112]
%onnx::Conv_789[FLOAT, 336x112x1x1]
%onnx::Conv_790[FLOAT, 336]
%onnx::Conv_792[FLOAT, 336x1x3x3]
%onnx::Conv_795[FLOAT, 112x336x1x1]
%onnx::Conv_798[FLOAT, 112x56x1x1]
%onnx::Conv_801[FLOAT, 112x1x3x3]
%onnx::Conv_804[FLOAT, 112x56x1x1]
%onnx::Conv_807[FLOAT, 112x112x1x1]
%onnx::Conv_810[FLOAT, 112x1x5x5]
%onnx::Conv_813[FLOAT, 112x112x1x1]
%onnx::Conv_816[FLOAT, 184x112x1x1]
%onnx::Conv_817[FLOAT, 184]
%onnx::Conv_819[FLOAT, 1104x184x1x1]
%onnx::Conv_820[FLOAT, 1104]
%onnx::Conv_822[FLOAT, 1104x1x5x5]
%onnx::Conv_825[FLOAT, 184x1104x1x1]
%onnx::Conv_828[FLOAT, 184x184x1x1]
%onnx::Conv_831[FLOAT, 184x1x5x5]
%onnx::Conv_834[FLOAT, 184x184x1x1]
%onnx::Conv_837[FLOAT, 552x184x1x1]
%onnx::Conv_838[FLOAT, 552]
%onnx::Conv_840[FLOAT, 552x1x3x3]
%onnx::Conv_843[FLOAT, 184x552x1x1]
%onnx::Conv_846[FLOAT, 552x184x1x1]
%onnx::Conv_849[FLOAT, 552x1x5x5]
%onnx::Conv_852[FLOAT, 352x552x1x1]
%onnx::Conv_853[FLOAT, 352]
%onnx::Conv_855[FLOAT, 1504x352x1x1]
%onnx::Conv_856[FLOAT, 1504]
) {
%onnx::Conv_850 = Identity(%onnx::Conv_838)
%onnx::Conv_847 = Identity(%onnx::Conv_838)
%onnx::Conv_844 = Identity(%onnx::Conv_817)
%onnx::Conv_841 = Identity(%onnx::Conv_838)
%onnx::Conv_835 = Identity(%onnx::Conv_817)
%onnx::Conv_832 = Identity(%onnx::Conv_817)
%onnx::Conv_829 = Identity(%onnx::Conv_817)
%onnx::Conv_826 = Identity(%onnx::Conv_817)
%onnx::Conv_823 = Identity(%onnx::Conv_820)
%onnx::Conv_814 = Identity(%onnx::Conv_787)
%onnx::Conv_811 = Identity(%onnx::Conv_787)
%onnx::Conv_808 = Identity(%onnx::Conv_787)
%onnx::Conv_805 = Identity(%onnx::Conv_787)
%onnx::Conv_802 = Identity(%onnx::Conv_787)
%onnx::Conv_799 = Identity(%onnx::Conv_787)
%onnx::Conv_796 = Identity(%onnx::Conv_787)
%onnx::Conv_793 = Identity(%onnx::Conv_790)
%onnx::Conv_784 = Identity(%onnx::Conv_781)
%onnx::Conv_778 = Identity(%onnx::Conv_760)
%onnx::Conv_775 = Identity(%onnx::Conv_760)
%onnx::Conv_772 = Identity(%onnx::Conv_760)
%onnx::Conv_769 = Identity(%onnx::Conv_760)
%onnx::Conv_766 = Identity(%onnx::Conv_763)
%onnx::Conv_757 = Identity(%onnx::Conv_724)
%onnx::Conv_754 = Identity(%onnx::Conv_724)
%onnx::Conv_751 = Identity(%onnx::Conv_724)
%onnx::Conv_748 = Identity(%onnx::Conv_673)
%onnx::Conv_745 = Identity(%onnx::Conv_673)
%onnx::Conv_742 = Identity(%onnx::Conv_724)
%onnx::Conv_739 = Identity(%onnx::Conv_673)
%onnx::Conv_736 = Identity(%onnx::Conv_673)
%onnx::Conv_733 = Identity(%onnx::Conv_724)
%onnx::Conv_730 = Identity(%onnx::Conv_724)
%onnx::Conv_727 = Identity(%onnx::Conv_724)
%onnx::Conv_721 = Identity(%onnx::Conv_700)
%onnx::Conv_718 = Identity(%onnx::Conv_700)
%onnx::Conv_715 = Identity(%onnx::Conv_688)
%onnx::Conv_712 = Identity(%onnx::Conv_700)
%onnx::Conv_709 = Identity(%onnx::Conv_700)
%onnx::Conv_706 = Identity(%onnx::Conv_688)
%onnx::Conv_703 = Identity(%onnx::Conv_700)
%onnx::Conv_697 = Identity(%onnx::Conv_688)
%onnx::Conv_694 = Identity(%onnx::Conv_688)
%onnx::Conv_691 = Identity(%onnx::Conv_688)
%onnx::Conv_685 = Identity(%onnx::Conv_682)
%onnx::Conv_679 = Identity(%onnx::Conv_670)
%onnx::Conv_676 = Identity(%onnx::Conv_673)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_669, %onnx::Conv_670)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%667 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %667
}
|
val_accuracy
| 0
| 80,056,320
| 2,028,340
|
{'zcp_synflow': 77.74701920854775, 'zcp_zen': 69.04659271240234, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.1703462302684784, 'zcp_flops': 80056320.0, 'zcp_grad_norm': 30.15534210205078, 'zcp_grasp': -0.06549453735351562, 'zcp_jacov': -16.05515192087561, 'zcp_l2_norm': 642.5196533203125, 'zcp_nwot': 218.34292945589746, 'zcp_params': 2028340.0, 'zcp_plain': -0.00010475059389136732, 'zcp_snip': 52.51728820800781, 'lat_1080ti_1': 0.7280925740903874, 'lat_1080ti_32': 0.7067075264247338, 'lat_1080ti_64': 0.7281317755164315, 'lat_2080ti_1': 0.6852581157519416, 'lat_2080ti_32': 0.7386635634516009, 'lat_2080ti_64': 0.7389549756121014, 'lat_essential_ph_1': 0.24528301886792453, 'lat_eyeriss': 0.6254286395393612, 'lat_fpga': 0.6218644554171324, 'lat_gold_6226': 0.4115405826497161, 'lat_gold_6240': 0.5429858191267707, 'lat_pixel2': 0.391304347826087, 'lat_pixel3': 0.6051997812148813, 'lat_raspi4': 0.6935680876673664, 'lat_samsung_a50': 0.25263157894736843, 'lat_samsung_s7': 0.1968503937007874, 'lat_silver_4114': 0.5506198332654123, 'lat_silver_4210r': 0.5762592409872057, 'lat_titan_rtx_1': 0.6372416591500067, 'lat_titan_rtx_32': 0.6840139793648703, 'lat_titan_rtx_64': 0.7450220093800696, 'lat_titanx_1': 0.3363116680359164, 'lat_titanx_32': 0.7251563749469893, 'lat_titanx_64': 0.7104715143528582, 'lat_titanxp_1': 0.6015858319549424, 'lat_titanxp_32': 0.7184267869975929, 'lat_titanxp_64': 0.72049682023467}
| |
FBNet_4678
|
FBNet
|
4678
|
4678
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_668[FLOAT, 16x3x3x3]
%onnx::Conv_669[FLOAT, 16]
%onnx::Conv_671[FLOAT, 16x16x1x1]
%onnx::Conv_674[FLOAT, 16x1x5x5]
%onnx::Conv_677[FLOAT, 16x16x1x1]
%onnx::Conv_680[FLOAT, 24x16x1x1]
%onnx::Conv_681[FLOAT, 24]
%onnx::Conv_683[FLOAT, 24x12x1x1]
%onnx::Conv_686[FLOAT, 24x1x5x5]
%onnx::Conv_689[FLOAT, 24x12x1x1]
%onnx::Conv_692[FLOAT, 24x12x1x1]
%onnx::Conv_695[FLOAT, 24x1x3x3]
%onnx::Conv_698[FLOAT, 24x12x1x1]
%onnx::Conv_701[FLOAT, 144x24x1x1]
%onnx::Conv_702[FLOAT, 144]
%onnx::Conv_704[FLOAT, 144x1x3x3]
%onnx::Conv_707[FLOAT, 24x144x1x1]
%onnx::Conv_710[FLOAT, 24x12x1x1]
%onnx::Conv_713[FLOAT, 24x1x3x3]
%onnx::Conv_716[FLOAT, 32x12x1x1]
%onnx::Conv_717[FLOAT, 32]
%onnx::Conv_719[FLOAT, 32x16x1x1]
%onnx::Conv_722[FLOAT, 32x1x3x3]
%onnx::Conv_725[FLOAT, 32x16x1x1]
%onnx::Conv_728[FLOAT, 32x32x1x1]
%onnx::Conv_731[FLOAT, 32x1x3x3]
%onnx::Conv_734[FLOAT, 32x32x1x1]
%onnx::Conv_737[FLOAT, 32x32x1x1]
%onnx::Conv_740[FLOAT, 32x1x5x5]
%onnx::Conv_743[FLOAT, 32x32x1x1]
%onnx::Conv_746[FLOAT, 32x32x1x1]
%onnx::Conv_749[FLOAT, 32x1x5x5]
%onnx::Conv_752[FLOAT, 64x32x1x1]
%onnx::Conv_753[FLOAT, 64]
%onnx::Conv_755[FLOAT, 384x64x1x1]
%onnx::Conv_756[FLOAT, 384]
%onnx::Conv_758[FLOAT, 384x1x5x5]
%onnx::Conv_761[FLOAT, 64x384x1x1]
%onnx::Conv_764[FLOAT, 64x64x1x1]
%onnx::Conv_767[FLOAT, 64x1x5x5]
%onnx::Conv_770[FLOAT, 64x64x1x1]
%onnx::Conv_773[FLOAT, 192x64x1x1]
%onnx::Conv_774[FLOAT, 192]
%onnx::Conv_776[FLOAT, 192x1x5x5]
%onnx::Conv_779[FLOAT, 64x192x1x1]
%onnx::Conv_782[FLOAT, 384x64x1x1]
%onnx::Conv_785[FLOAT, 384x1x5x5]
%onnx::Conv_788[FLOAT, 112x384x1x1]
%onnx::Conv_789[FLOAT, 112]
%onnx::Conv_791[FLOAT, 336x112x1x1]
%onnx::Conv_792[FLOAT, 336]
%onnx::Conv_794[FLOAT, 336x1x3x3]
%onnx::Conv_797[FLOAT, 112x336x1x1]
%onnx::Conv_800[FLOAT, 112x112x1x1]
%onnx::Conv_803[FLOAT, 112x1x3x3]
%onnx::Conv_806[FLOAT, 112x112x1x1]
%onnx::Conv_809[FLOAT, 672x112x1x1]
%onnx::Conv_810[FLOAT, 672]
%onnx::Conv_812[FLOAT, 672x1x5x5]
%onnx::Conv_815[FLOAT, 184x672x1x1]
%onnx::Conv_816[FLOAT, 184]
%onnx::Conv_818[FLOAT, 552x184x1x1]
%onnx::Conv_819[FLOAT, 552]
%onnx::Conv_821[FLOAT, 552x1x3x3]
%onnx::Conv_824[FLOAT, 184x552x1x1]
%onnx::Conv_827[FLOAT, 552x184x1x1]
%onnx::Conv_830[FLOAT, 552x1x5x5]
%onnx::Conv_833[FLOAT, 184x552x1x1]
%onnx::Conv_836[FLOAT, 184x92x1x1]
%onnx::Conv_839[FLOAT, 184x1x5x5]
%onnx::Conv_842[FLOAT, 184x92x1x1]
%onnx::Conv_845[FLOAT, 184x184x1x1]
%onnx::Conv_848[FLOAT, 184x1x5x5]
%onnx::Conv_851[FLOAT, 352x184x1x1]
%onnx::Conv_852[FLOAT, 352]
%onnx::Conv_854[FLOAT, 1504x352x1x1]
%onnx::Conv_855[FLOAT, 1504]
) {
%onnx::Conv_849 = Identity(%onnx::Conv_816)
%onnx::Conv_846 = Identity(%onnx::Conv_816)
%onnx::Conv_843 = Identity(%onnx::Conv_816)
%onnx::Conv_840 = Identity(%onnx::Conv_816)
%onnx::Conv_837 = Identity(%onnx::Conv_816)
%onnx::Conv_834 = Identity(%onnx::Conv_816)
%onnx::Conv_831 = Identity(%onnx::Conv_819)
%onnx::Conv_828 = Identity(%onnx::Conv_819)
%onnx::Conv_825 = Identity(%onnx::Conv_816)
%onnx::Conv_822 = Identity(%onnx::Conv_819)
%onnx::Conv_813 = Identity(%onnx::Conv_810)
%onnx::Conv_807 = Identity(%onnx::Conv_789)
%onnx::Conv_804 = Identity(%onnx::Conv_789)
%onnx::Conv_801 = Identity(%onnx::Conv_789)
%onnx::Conv_798 = Identity(%onnx::Conv_789)
%onnx::Conv_795 = Identity(%onnx::Conv_792)
%onnx::Conv_786 = Identity(%onnx::Conv_756)
%onnx::Conv_783 = Identity(%onnx::Conv_756)
%onnx::Conv_780 = Identity(%onnx::Conv_753)
%onnx::Conv_777 = Identity(%onnx::Conv_774)
%onnx::Conv_771 = Identity(%onnx::Conv_753)
%onnx::Conv_768 = Identity(%onnx::Conv_753)
%onnx::Conv_765 = Identity(%onnx::Conv_753)
%onnx::Conv_762 = Identity(%onnx::Conv_753)
%onnx::Conv_759 = Identity(%onnx::Conv_756)
%onnx::Conv_750 = Identity(%onnx::Conv_717)
%onnx::Conv_747 = Identity(%onnx::Conv_717)
%onnx::Conv_744 = Identity(%onnx::Conv_717)
%onnx::Conv_741 = Identity(%onnx::Conv_717)
%onnx::Conv_738 = Identity(%onnx::Conv_717)
%onnx::Conv_735 = Identity(%onnx::Conv_717)
%onnx::Conv_732 = Identity(%onnx::Conv_717)
%onnx::Conv_729 = Identity(%onnx::Conv_717)
%onnx::Conv_726 = Identity(%onnx::Conv_717)
%onnx::Conv_723 = Identity(%onnx::Conv_717)
%onnx::Conv_720 = Identity(%onnx::Conv_717)
%onnx::Conv_714 = Identity(%onnx::Conv_681)
%onnx::Conv_711 = Identity(%onnx::Conv_681)
%onnx::Conv_708 = Identity(%onnx::Conv_681)
%onnx::Conv_705 = Identity(%onnx::Conv_702)
%onnx::Conv_699 = Identity(%onnx::Conv_681)
%onnx::Conv_696 = Identity(%onnx::Conv_681)
%onnx::Conv_693 = Identity(%onnx::Conv_681)
%onnx::Conv_690 = Identity(%onnx::Conv_681)
%onnx::Conv_687 = Identity(%onnx::Conv_681)
%onnx::Conv_684 = Identity(%onnx::Conv_681)
%onnx::Conv_678 = Identity(%onnx::Conv_669)
%onnx::Conv_675 = Identity(%onnx::Conv_669)
%onnx::Conv_672 = Identity(%onnx::Conv_669)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_668, %onnx::Conv_669)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.1/relu/Relu_output_0 = Relu(%/cells.1/conv/Conv_output_0)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/relu/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/relu/Relu_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_803, %onnx::Conv_804)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_806, %onnx::Conv_807)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_809, %onnx::Conv_810)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_812, %onnx::Conv_813)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_815, %onnx::Conv_816)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_818, %onnx::Conv_819)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_821, %onnx::Conv_822)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_824, %onnx::Conv_825)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_827, %onnx::Conv_828)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_830, %onnx::Conv_831)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_833, %onnx::Conv_834)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_836, %onnx::Conv_837)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_839, %onnx::Conv_840)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_842, %onnx::Conv_843)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_845, %onnx::Conv_846)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_848, %onnx::Conv_849)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_851, %onnx::Conv_852)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_854, %onnx::Conv_855)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%666 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %666
}
|
val_accuracy
| 0
| 59,945,856
| 1,787,772
|
{'zcp_synflow': 78.07526469564506, 'zcp_zen': 67.55355834960938, 'zcp_epe_nas': 19.644875916588823, 'zcp_fisher': 0.10412926226854324, 'zcp_flops': 59945856.0, 'zcp_grad_norm': 21.277803421020508, 'zcp_grasp': -0.055896759033203125, 'zcp_jacov': -16.057909483997236, 'zcp_l2_norm': 611.2105712890625, 'zcp_nwot': 209.4384318814301, 'zcp_params': 1787772.0, 'zcp_plain': -0.0006725740968249738, 'zcp_snip': 38.86490249633789, 'lat_1080ti_1': 0.583970200706176, 'lat_1080ti_32': 0.5786474989121444, 'lat_1080ti_64': 0.37625088661770506, 'lat_2080ti_1': 0.6687761872776968, 'lat_2080ti_32': 0.5791765116877217, 'lat_2080ti_64': 0.39508381288075367, 'lat_essential_ph_1': 0.2641509433962264, 'lat_eyeriss': 0.33603506133900596, 'lat_fpga': 0.3441059183249387, 'lat_gold_6226': 0.31733447915264823, 'lat_gold_6240': 0.46460378175112577, 'lat_pixel2': 0.2826086956521739, 'lat_pixel3': 0.3345421811462228, 'lat_raspi4': 0.3362462982377357, 'lat_samsung_a50': 0.17894736842105263, 'lat_samsung_s7': 0.15748031496062992, 'lat_silver_4114': 0.5506172385867193, 'lat_silver_4210r': 0.5571541020028992, 'lat_titan_rtx_1': 0.6304484922897257, 'lat_titan_rtx_32': 0.5674152739951325, 'lat_titan_rtx_64': 0.43812024996520577, 'lat_titanx_1': 0.3271679267082263, 'lat_titanx_32': 0.49089211513821196, 'lat_titanx_64': 0.3762305196251178, 'lat_titanxp_1': 0.581984086912795, 'lat_titanxp_32': 0.5531083288049309, 'lat_titanxp_64': 0.38703729244130847}
| |
FBNet_1797
|
FBNet
|
1797
|
1797
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_678[FLOAT, 16x3x3x3]
%onnx::Conv_679[FLOAT, 16]
%onnx::Conv_681[FLOAT, 16x8x1x1]
%onnx::Conv_684[FLOAT, 16x1x3x3]
%onnx::Conv_687[FLOAT, 16x8x1x1]
%onnx::Conv_690[FLOAT, 16x16x1x1]
%onnx::Conv_693[FLOAT, 16x1x5x5]
%onnx::Conv_696[FLOAT, 24x16x1x1]
%onnx::Conv_697[FLOAT, 24]
%onnx::Conv_699[FLOAT, 24x24x1x1]
%onnx::Conv_702[FLOAT, 24x1x3x3]
%onnx::Conv_705[FLOAT, 24x24x1x1]
%onnx::Conv_708[FLOAT, 24x24x1x1]
%onnx::Conv_711[FLOAT, 24x1x3x3]
%onnx::Conv_714[FLOAT, 24x24x1x1]
%onnx::Conv_717[FLOAT, 72x24x1x1]
%onnx::Conv_718[FLOAT, 72]
%onnx::Conv_720[FLOAT, 72x1x5x5]
%onnx::Conv_723[FLOAT, 24x72x1x1]
%onnx::Conv_726[FLOAT, 24x24x1x1]
%onnx::Conv_729[FLOAT, 24x1x5x5]
%onnx::Conv_732[FLOAT, 32x24x1x1]
%onnx::Conv_733[FLOAT, 32]
%onnx::Conv_735[FLOAT, 96x32x1x1]
%onnx::Conv_736[FLOAT, 96]
%onnx::Conv_738[FLOAT, 96x1x5x5]
%onnx::Conv_741[FLOAT, 32x96x1x1]
%onnx::Conv_744[FLOAT, 32x16x1x1]
%onnx::Conv_747[FLOAT, 32x1x3x3]
%onnx::Conv_750[FLOAT, 32x16x1x1]
%onnx::Conv_753[FLOAT, 32x16x1x1]
%onnx::Conv_756[FLOAT, 32x1x3x3]
%onnx::Conv_759[FLOAT, 64x16x1x1]
%onnx::Conv_760[FLOAT, 64]
%onnx::Conv_762[FLOAT, 192x64x1x1]
%onnx::Conv_763[FLOAT, 192]
%onnx::Conv_765[FLOAT, 192x1x5x5]
%onnx::Conv_768[FLOAT, 64x192x1x1]
%onnx::Conv_771[FLOAT, 64x32x1x1]
%onnx::Conv_774[FLOAT, 64x1x3x3]
%onnx::Conv_777[FLOAT, 64x32x1x1]
%onnx::Conv_780[FLOAT, 64x64x1x1]
%onnx::Conv_783[FLOAT, 64x1x3x3]
%onnx::Conv_786[FLOAT, 64x64x1x1]
%onnx::Conv_789[FLOAT, 192x64x1x1]
%onnx::Conv_792[FLOAT, 192x1x3x3]
%onnx::Conv_795[FLOAT, 112x192x1x1]
%onnx::Conv_796[FLOAT, 112]
%onnx::Conv_798[FLOAT, 672x112x1x1]
%onnx::Conv_799[FLOAT, 672]
%onnx::Conv_801[FLOAT, 672x1x3x3]
%onnx::Conv_804[FLOAT, 112x672x1x1]
%onnx::Conv_807[FLOAT, 112x56x1x1]
%onnx::Conv_810[FLOAT, 112x1x3x3]
%onnx::Conv_813[FLOAT, 112x56x1x1]
%onnx::Conv_816[FLOAT, 336x112x1x1]
%onnx::Conv_817[FLOAT, 336]
%onnx::Conv_819[FLOAT, 336x1x3x3]
%onnx::Conv_822[FLOAT, 184x336x1x1]
%onnx::Conv_823[FLOAT, 184]
%onnx::Conv_825[FLOAT, 184x184x1x1]
%onnx::Conv_828[FLOAT, 184x1x3x3]
%onnx::Conv_831[FLOAT, 184x184x1x1]
%onnx::Conv_834[FLOAT, 184x92x1x1]
%onnx::Conv_837[FLOAT, 184x1x3x3]
%onnx::Conv_840[FLOAT, 184x92x1x1]
%onnx::Conv_843[FLOAT, 552x184x1x1]
%onnx::Conv_844[FLOAT, 552]
%onnx::Conv_846[FLOAT, 552x1x5x5]
%onnx::Conv_849[FLOAT, 184x552x1x1]
%onnx::Conv_852[FLOAT, 184x184x1x1]
%onnx::Conv_855[FLOAT, 184x1x3x3]
%onnx::Conv_858[FLOAT, 352x184x1x1]
%onnx::Conv_859[FLOAT, 352]
%onnx::Conv_861[FLOAT, 1504x352x1x1]
%onnx::Conv_862[FLOAT, 1504]
) {
%onnx::Conv_856 = Identity(%onnx::Conv_823)
%onnx::Conv_853 = Identity(%onnx::Conv_823)
%onnx::Conv_850 = Identity(%onnx::Conv_823)
%onnx::Conv_847 = Identity(%onnx::Conv_844)
%onnx::Conv_841 = Identity(%onnx::Conv_823)
%onnx::Conv_838 = Identity(%onnx::Conv_823)
%onnx::Conv_835 = Identity(%onnx::Conv_823)
%onnx::Conv_832 = Identity(%onnx::Conv_823)
%onnx::Conv_829 = Identity(%onnx::Conv_823)
%onnx::Conv_826 = Identity(%onnx::Conv_823)
%onnx::Conv_820 = Identity(%onnx::Conv_817)
%onnx::Conv_814 = Identity(%onnx::Conv_796)
%onnx::Conv_811 = Identity(%onnx::Conv_796)
%onnx::Conv_808 = Identity(%onnx::Conv_796)
%onnx::Conv_805 = Identity(%onnx::Conv_796)
%onnx::Conv_802 = Identity(%onnx::Conv_799)
%onnx::Conv_793 = Identity(%onnx::Conv_763)
%onnx::Conv_790 = Identity(%onnx::Conv_763)
%onnx::Conv_787 = Identity(%onnx::Conv_760)
%onnx::Conv_784 = Identity(%onnx::Conv_760)
%onnx::Conv_781 = Identity(%onnx::Conv_760)
%onnx::Conv_778 = Identity(%onnx::Conv_760)
%onnx::Conv_775 = Identity(%onnx::Conv_760)
%onnx::Conv_772 = Identity(%onnx::Conv_760)
%onnx::Conv_769 = Identity(%onnx::Conv_760)
%onnx::Conv_766 = Identity(%onnx::Conv_763)
%onnx::Conv_757 = Identity(%onnx::Conv_733)
%onnx::Conv_754 = Identity(%onnx::Conv_733)
%onnx::Conv_751 = Identity(%onnx::Conv_733)
%onnx::Conv_748 = Identity(%onnx::Conv_733)
%onnx::Conv_745 = Identity(%onnx::Conv_733)
%onnx::Conv_742 = Identity(%onnx::Conv_733)
%onnx::Conv_739 = Identity(%onnx::Conv_736)
%onnx::Conv_730 = Identity(%onnx::Conv_697)
%onnx::Conv_727 = Identity(%onnx::Conv_697)
%onnx::Conv_724 = Identity(%onnx::Conv_697)
%onnx::Conv_721 = Identity(%onnx::Conv_718)
%onnx::Conv_715 = Identity(%onnx::Conv_697)
%onnx::Conv_712 = Identity(%onnx::Conv_697)
%onnx::Conv_709 = Identity(%onnx::Conv_697)
%onnx::Conv_706 = Identity(%onnx::Conv_697)
%onnx::Conv_703 = Identity(%onnx::Conv_697)
%onnx::Conv_700 = Identity(%onnx::Conv_697)
%onnx::Conv_694 = Identity(%onnx::Conv_679)
%onnx::Conv_691 = Identity(%onnx::Conv_679)
%onnx::Conv_688 = Identity(%onnx::Conv_679)
%onnx::Conv_685 = Identity(%onnx::Conv_679)
%onnx::Conv_682 = Identity(%onnx::Conv_679)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_678, %onnx::Conv_679)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_858, %onnx::Conv_859)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_861, %onnx::Conv_862)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%676 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %676
}
|
val_accuracy
| 0
| 50,630,016
| 1,495,540
|
{'zcp_synflow': 72.85627383744372, 'zcp_zen': 62.812496185302734, 'zcp_epe_nas': 10.903739574597507, 'zcp_fisher': 0.09107808768749237, 'zcp_flops': 50630016.0, 'zcp_grad_norm': 19.121543884277344, 'zcp_grasp': -0.08646011352539062, 'zcp_jacov': -16.057528359360873, 'zcp_l2_norm': 551.3997192382812, 'zcp_nwot': 205.60930397619418, 'zcp_params': 1495540.0, 'zcp_plain': -0.004583777394145727, 'zcp_snip': 33.79859161376953, 'lat_1080ti_1': 0.5787420012673878, 'lat_1080ti_32': 0.49278384662969255, 'lat_1080ti_64': 0.3259187651517793, 'lat_2080ti_1': 0.6384497920517309, 'lat_2080ti_32': 0.5049358959459971, 'lat_2080ti_64': 0.31570032579407653, 'lat_essential_ph_1': 0.3018867924528302, 'lat_eyeriss': 0.19239726404195612, 'lat_fpga': 0.24114626109500387, 'lat_gold_6226': 0.18675154191946916, 'lat_gold_6240': 0.3696807474254218, 'lat_pixel2': 0.2608695652173913, 'lat_pixel3': 0.2156336551300304, 'lat_raspi4': 0.23349713362726582, 'lat_samsung_a50': 0.09473684210526316, 'lat_samsung_s7': 0.09448818897637795, 'lat_silver_4114': 0.6043371921629379, 'lat_silver_4210r': 0.45010146200781925, 'lat_titan_rtx_1': 0.6065542728606511, 'lat_titan_rtx_32': 0.5110148501848754, 'lat_titan_rtx_64': 0.3669856485863455, 'lat_titanx_1': 0.3209451832702501, 'lat_titanx_32': 0.41981141952880496, 'lat_titanx_64': 0.2742022316335454, 'lat_titanxp_1': 0.5715439487467414, 'lat_titanxp_32': 0.4757489643093599, 'lat_titanxp_64': 0.31210957917526505}
| |
FBNet_1326
|
FBNet
|
1326
|
1326
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_579[FLOAT, 16x3x3x3]
%onnx::Conv_580[FLOAT, 16]
%onnx::Conv_582[FLOAT, 96x16x1x1]
%onnx::Conv_583[FLOAT, 96]
%onnx::Conv_585[FLOAT, 96x1x5x5]
%onnx::Conv_588[FLOAT, 24x96x1x1]
%onnx::Conv_589[FLOAT, 24]
%onnx::Conv_591[FLOAT, 24x12x1x1]
%onnx::Conv_594[FLOAT, 24x1x5x5]
%onnx::Conv_597[FLOAT, 24x12x1x1]
%onnx::Conv_600[FLOAT, 72x24x1x1]
%onnx::Conv_601[FLOAT, 72]
%onnx::Conv_603[FLOAT, 72x1x5x5]
%onnx::Conv_606[FLOAT, 24x72x1x1]
%onnx::Conv_609[FLOAT, 24x12x1x1]
%onnx::Conv_612[FLOAT, 24x1x3x3]
%onnx::Conv_615[FLOAT, 24x12x1x1]
%onnx::Conv_618[FLOAT, 32x24x1x1]
%onnx::Conv_619[FLOAT, 32]
%onnx::Conv_621[FLOAT, 192x32x1x1]
%onnx::Conv_622[FLOAT, 192]
%onnx::Conv_624[FLOAT, 192x1x3x3]
%onnx::Conv_627[FLOAT, 32x192x1x1]
%onnx::Conv_630[FLOAT, 96x32x1x1]
%onnx::Conv_633[FLOAT, 96x1x3x3]
%onnx::Conv_636[FLOAT, 32x96x1x1]
%onnx::Conv_639[FLOAT, 192x32x1x1]
%onnx::Conv_642[FLOAT, 192x1x5x5]
%onnx::Conv_645[FLOAT, 32x192x1x1]
%onnx::Conv_648[FLOAT, 64x32x1x1]
%onnx::Conv_649[FLOAT, 64]
%onnx::Conv_651[FLOAT, 192x64x1x1]
%onnx::Conv_654[FLOAT, 192x1x5x5]
%onnx::Conv_657[FLOAT, 64x192x1x1]
%onnx::Conv_660[FLOAT, 64x32x1x1]
%onnx::Conv_663[FLOAT, 64x1x3x3]
%onnx::Conv_666[FLOAT, 112x32x1x1]
%onnx::Conv_667[FLOAT, 112]
%onnx::Conv_669[FLOAT, 112x56x1x1]
%onnx::Conv_672[FLOAT, 112x1x5x5]
%onnx::Conv_675[FLOAT, 112x56x1x1]
%onnx::Conv_678[FLOAT, 112x112x1x1]
%onnx::Conv_681[FLOAT, 112x1x5x5]
%onnx::Conv_684[FLOAT, 112x112x1x1]
%onnx::Conv_687[FLOAT, 112x112x1x1]
%onnx::Conv_690[FLOAT, 112x1x3x3]
%onnx::Conv_693[FLOAT, 112x112x1x1]
%onnx::Conv_696[FLOAT, 336x112x1x1]
%onnx::Conv_697[FLOAT, 336]
%onnx::Conv_699[FLOAT, 336x1x3x3]
%onnx::Conv_702[FLOAT, 184x336x1x1]
%onnx::Conv_703[FLOAT, 184]
%onnx::Conv_705[FLOAT, 552x184x1x1]
%onnx::Conv_706[FLOAT, 552]
%onnx::Conv_708[FLOAT, 552x1x5x5]
%onnx::Conv_711[FLOAT, 184x552x1x1]
%onnx::Conv_714[FLOAT, 1104x184x1x1]
%onnx::Conv_715[FLOAT, 1104]
%onnx::Conv_717[FLOAT, 1104x1x5x5]
%onnx::Conv_720[FLOAT, 184x1104x1x1]
%onnx::Conv_723[FLOAT, 1104x184x1x1]
%onnx::Conv_726[FLOAT, 1104x1x5x5]
%onnx::Conv_729[FLOAT, 184x1104x1x1]
%onnx::Conv_732[FLOAT, 552x184x1x1]
%onnx::Conv_735[FLOAT, 552x1x5x5]
%onnx::Conv_738[FLOAT, 352x552x1x1]
%onnx::Conv_739[FLOAT, 352]
%onnx::Conv_741[FLOAT, 1504x352x1x1]
%onnx::Conv_742[FLOAT, 1504]
) {
%onnx::Conv_736 = Identity(%onnx::Conv_706)
%onnx::Conv_733 = Identity(%onnx::Conv_706)
%onnx::Conv_730 = Identity(%onnx::Conv_703)
%onnx::Conv_727 = Identity(%onnx::Conv_715)
%onnx::Conv_724 = Identity(%onnx::Conv_715)
%onnx::Conv_721 = Identity(%onnx::Conv_703)
%onnx::Conv_718 = Identity(%onnx::Conv_715)
%onnx::Conv_712 = Identity(%onnx::Conv_703)
%onnx::Conv_709 = Identity(%onnx::Conv_706)
%onnx::Conv_700 = Identity(%onnx::Conv_697)
%onnx::Conv_694 = Identity(%onnx::Conv_667)
%onnx::Conv_691 = Identity(%onnx::Conv_667)
%onnx::Conv_688 = Identity(%onnx::Conv_667)
%onnx::Conv_685 = Identity(%onnx::Conv_667)
%onnx::Conv_682 = Identity(%onnx::Conv_667)
%onnx::Conv_679 = Identity(%onnx::Conv_667)
%onnx::Conv_676 = Identity(%onnx::Conv_667)
%onnx::Conv_673 = Identity(%onnx::Conv_667)
%onnx::Conv_670 = Identity(%onnx::Conv_667)
%onnx::Conv_664 = Identity(%onnx::Conv_649)
%onnx::Conv_661 = Identity(%onnx::Conv_649)
%onnx::Conv_658 = Identity(%onnx::Conv_649)
%onnx::Conv_655 = Identity(%onnx::Conv_622)
%onnx::Conv_652 = Identity(%onnx::Conv_622)
%onnx::Conv_646 = Identity(%onnx::Conv_619)
%onnx::Conv_643 = Identity(%onnx::Conv_622)
%onnx::Conv_640 = Identity(%onnx::Conv_622)
%onnx::Conv_637 = Identity(%onnx::Conv_619)
%onnx::Conv_634 = Identity(%onnx::Conv_583)
%onnx::Conv_631 = Identity(%onnx::Conv_583)
%onnx::Conv_628 = Identity(%onnx::Conv_619)
%onnx::Conv_625 = Identity(%onnx::Conv_622)
%onnx::Conv_616 = Identity(%onnx::Conv_589)
%onnx::Conv_613 = Identity(%onnx::Conv_589)
%onnx::Conv_610 = Identity(%onnx::Conv_589)
%onnx::Conv_607 = Identity(%onnx::Conv_589)
%onnx::Conv_604 = Identity(%onnx::Conv_601)
%onnx::Conv_598 = Identity(%onnx::Conv_589)
%onnx::Conv_595 = Identity(%onnx::Conv_589)
%onnx::Conv_592 = Identity(%onnx::Conv_589)
%onnx::Conv_586 = Identity(%onnx::Conv_583)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_579, %onnx::Conv_580)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_582, %onnx::Conv_583)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_585, %onnx::Conv_586)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_588, %onnx::Conv_589)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_591, %onnx::Conv_592)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_594, %onnx::Conv_595)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_597, %onnx::Conv_598)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_600, %onnx::Conv_601)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_603, %onnx::Conv_604)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_606, %onnx::Conv_607)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_609, %onnx::Conv_610)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_612, %onnx::Conv_613)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_615, %onnx::Conv_616)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.4/Add_output_0, %onnx::Conv_618, %onnx::Conv_619)
%/cells.5/relu/Relu_output_0 = Relu(%/cells.5/conv/Conv_output_0)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/relu/Relu_output_0, %onnx::Conv_621, %onnx::Conv_622)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_624, %onnx::Conv_625)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_627, %onnx::Conv_628)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/relu/Relu_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_630, %onnx::Conv_631)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.8/Add_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.9/relu/Relu_output_0 = Relu(%/cells.9/conv/Conv_output_0)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/relu/Relu_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/relu/Relu_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%577 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %577
}
|
val_accuracy
| 0
| 68,167,040
| 2,363,036
|
{'zcp_synflow': 70.93587696592316, 'zcp_zen': 63.479862213134766, 'zcp_epe_nas': 6.854940869244546, 'zcp_fisher': 0.07163454592227936, 'zcp_flops': 68167040.0, 'zcp_grad_norm': 19.047395706176758, 'zcp_grasp': -0.017543792724609375, 'zcp_jacov': -16.070719326781656, 'zcp_l2_norm': 609.3319091796875, 'zcp_nwot': 210.74908171187607, 'zcp_params': 2363036.0, 'zcp_plain': 0.008945860899984837, 'zcp_snip': 32.305908203125, 'lat_1080ti_1': 0.30223627022390703, 'lat_1080ti_32': 0.36422066239370837, 'lat_1080ti_64': 0.3566648579616734, 'lat_2080ti_1': 0.3595983717726517, 'lat_2080ti_32': 0.32974077506233374, 'lat_2080ti_64': 0.33356333307207153, 'lat_essential_ph_1': 0.33962264150943394, 'lat_eyeriss': 0.5048227678653291, 'lat_fpga': 0.4631947674130952, 'lat_gold_6226': 0.588031563655933, 'lat_gold_6240': 0.5250818126535531, 'lat_pixel2': 0.30434782608695654, 'lat_pixel3': 0.49513295173986877, 'lat_raspi4': 0.5460462318759379, 'lat_samsung_a50': 0.2, 'lat_samsung_s7': 0.1732283464566929, 'lat_silver_4114': 0.5358744164004852, 'lat_silver_4210r': 0.521009914973791, 'lat_titan_rtx_1': 0.35007960558285345, 'lat_titan_rtx_32': 0.325957641274243, 'lat_titan_rtx_64': 0.3318296495427798, 'lat_titanx_1': 0.18311428713179684, 'lat_titanx_32': 0.3228729481484903, 'lat_titanx_64': 0.3497099526221658, 'lat_titanxp_1': 0.33533412534041673, 'lat_titanxp_32': 0.33627741107533143, 'lat_titanxp_64': 0.3445296114963725}
| |
FBNet_4844
|
FBNet
|
4844
|
4844
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_487[FLOAT, 16x3x3x3]
%onnx::Conv_488[FLOAT, 16]
%onnx::Conv_490[FLOAT, 16x8x1x1]
%onnx::Conv_493[FLOAT, 16x1x3x3]
%onnx::Conv_496[FLOAT, 16x8x1x1]
%onnx::Conv_499[FLOAT, 96x16x1x1]
%onnx::Conv_500[FLOAT, 96]
%onnx::Conv_502[FLOAT, 96x1x3x3]
%onnx::Conv_505[FLOAT, 24x96x1x1]
%onnx::Conv_506[FLOAT, 24]
%onnx::Conv_508[FLOAT, 72x24x1x1]
%onnx::Conv_509[FLOAT, 72]
%onnx::Conv_511[FLOAT, 72x1x3x3]
%onnx::Conv_514[FLOAT, 24x72x1x1]
%onnx::Conv_517[FLOAT, 24x24x1x1]
%onnx::Conv_520[FLOAT, 24x1x3x3]
%onnx::Conv_523[FLOAT, 24x24x1x1]
%onnx::Conv_526[FLOAT, 24x24x1x1]
%onnx::Conv_529[FLOAT, 24x1x5x5]
%onnx::Conv_532[FLOAT, 32x24x1x1]
%onnx::Conv_533[FLOAT, 32]
%onnx::Conv_535[FLOAT, 96x32x1x1]
%onnx::Conv_538[FLOAT, 96x1x3x3]
%onnx::Conv_541[FLOAT, 32x96x1x1]
%onnx::Conv_544[FLOAT, 96x32x1x1]
%onnx::Conv_547[FLOAT, 96x1x3x3]
%onnx::Conv_550[FLOAT, 32x96x1x1]
%onnx::Conv_553[FLOAT, 32x32x1x1]
%onnx::Conv_556[FLOAT, 32x1x5x5]
%onnx::Conv_559[FLOAT, 64x32x1x1]
%onnx::Conv_560[FLOAT, 64]
%onnx::Conv_562[FLOAT, 64x32x1x1]
%onnx::Conv_565[FLOAT, 64x1x3x3]
%onnx::Conv_568[FLOAT, 112x32x1x1]
%onnx::Conv_569[FLOAT, 112]
%onnx::Conv_571[FLOAT, 672x112x1x1]
%onnx::Conv_572[FLOAT, 672]
%onnx::Conv_574[FLOAT, 672x1x3x3]
%onnx::Conv_577[FLOAT, 112x672x1x1]
%onnx::Conv_580[FLOAT, 336x112x1x1]
%onnx::Conv_581[FLOAT, 336]
%onnx::Conv_583[FLOAT, 336x1x3x3]
%onnx::Conv_586[FLOAT, 112x336x1x1]
%onnx::Conv_589[FLOAT, 112x56x1x1]
%onnx::Conv_592[FLOAT, 112x1x5x5]
%onnx::Conv_595[FLOAT, 184x56x1x1]
%onnx::Conv_596[FLOAT, 184]
%onnx::Conv_598[FLOAT, 552x184x1x1]
%onnx::Conv_599[FLOAT, 552]
%onnx::Conv_601[FLOAT, 552x1x5x5]
%onnx::Conv_604[FLOAT, 184x552x1x1]
%onnx::Conv_607[FLOAT, 1104x184x1x1]
%onnx::Conv_608[FLOAT, 1104]
%onnx::Conv_610[FLOAT, 1104x1x5x5]
%onnx::Conv_613[FLOAT, 184x1104x1x1]
%onnx::Conv_616[FLOAT, 1104x184x1x1]
%onnx::Conv_619[FLOAT, 1104x1x5x5]
%onnx::Conv_622[FLOAT, 352x1104x1x1]
%onnx::Conv_623[FLOAT, 352]
%onnx::Conv_625[FLOAT, 1504x352x1x1]
%onnx::Conv_626[FLOAT, 1504]
) {
%onnx::Conv_620 = Identity(%onnx::Conv_608)
%onnx::Conv_617 = Identity(%onnx::Conv_608)
%onnx::Conv_614 = Identity(%onnx::Conv_596)
%onnx::Conv_611 = Identity(%onnx::Conv_608)
%onnx::Conv_605 = Identity(%onnx::Conv_596)
%onnx::Conv_602 = Identity(%onnx::Conv_599)
%onnx::Conv_593 = Identity(%onnx::Conv_569)
%onnx::Conv_590 = Identity(%onnx::Conv_569)
%onnx::Conv_587 = Identity(%onnx::Conv_569)
%onnx::Conv_584 = Identity(%onnx::Conv_581)
%onnx::Conv_578 = Identity(%onnx::Conv_569)
%onnx::Conv_575 = Identity(%onnx::Conv_572)
%onnx::Conv_566 = Identity(%onnx::Conv_560)
%onnx::Conv_563 = Identity(%onnx::Conv_560)
%onnx::Conv_557 = Identity(%onnx::Conv_533)
%onnx::Conv_554 = Identity(%onnx::Conv_533)
%onnx::Conv_551 = Identity(%onnx::Conv_533)
%onnx::Conv_548 = Identity(%onnx::Conv_500)
%onnx::Conv_545 = Identity(%onnx::Conv_500)
%onnx::Conv_542 = Identity(%onnx::Conv_533)
%onnx::Conv_539 = Identity(%onnx::Conv_500)
%onnx::Conv_536 = Identity(%onnx::Conv_500)
%onnx::Conv_530 = Identity(%onnx::Conv_506)
%onnx::Conv_527 = Identity(%onnx::Conv_506)
%onnx::Conv_524 = Identity(%onnx::Conv_506)
%onnx::Conv_521 = Identity(%onnx::Conv_506)
%onnx::Conv_518 = Identity(%onnx::Conv_506)
%onnx::Conv_515 = Identity(%onnx::Conv_506)
%onnx::Conv_512 = Identity(%onnx::Conv_509)
%onnx::Conv_503 = Identity(%onnx::Conv_500)
%onnx::Conv_497 = Identity(%onnx::Conv_488)
%onnx::Conv_494 = Identity(%onnx::Conv_488)
%onnx::Conv_491 = Identity(%onnx::Conv_488)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_487, %onnx::Conv_488)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_490, %onnx::Conv_491)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_493, %onnx::Conv_494)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_496, %onnx::Conv_497)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_499, %onnx::Conv_500)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_502, %onnx::Conv_503)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_505, %onnx::Conv_506)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_508, %onnx::Conv_509)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_511, %onnx::Conv_512)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_514, %onnx::Conv_515)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_517, %onnx::Conv_518)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_520, %onnx::Conv_521)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_523, %onnx::Conv_524)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_526, %onnx::Conv_527)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_529, %onnx::Conv_530)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_532, %onnx::Conv_533)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_535, %onnx::Conv_536)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_538, %onnx::Conv_539)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_541, %onnx::Conv_542)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_544, %onnx::Conv_545)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_547, %onnx::Conv_548)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_550, %onnx::Conv_551)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_553, %onnx::Conv_554)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_556, %onnx::Conv_557)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_559, %onnx::Conv_560)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_562, %onnx::Conv_563)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_565, %onnx::Conv_566)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_568, %onnx::Conv_569)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_571, %onnx::Conv_572)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_574, %onnx::Conv_575)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_577, %onnx::Conv_578)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_580, %onnx::Conv_581)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_583, %onnx::Conv_584)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_586, %onnx::Conv_587)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_589, %onnx::Conv_590)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_592, %onnx::Conv_593)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_595, %onnx::Conv_596)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_598, %onnx::Conv_599)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_601, %onnx::Conv_602)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_604, %onnx::Conv_605)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_607, %onnx::Conv_608)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_610, %onnx::Conv_611)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_613, %onnx::Conv_614)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_616, %onnx::Conv_617)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_619, %onnx::Conv_620)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_622, %onnx::Conv_623)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_625, %onnx::Conv_626)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%485 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %485
}
|
val_accuracy
| 0
| 63,667,072
| 2,264,964
|
{'zcp_synflow': 58.31856557562351, 'zcp_zen': 51.543453216552734, 'zcp_epe_nas': 7.773143183074044, 'zcp_fisher': 0.03094910830259323, 'zcp_flops': 63667072.0, 'zcp_grad_norm': 12.92577075958252, 'zcp_grasp': 0.005592823028564453, 'zcp_jacov': -16.05293051363445, 'zcp_l2_norm': 499.95623779296875, 'zcp_nwot': 208.77257837853196, 'zcp_params': 2264964.0, 'zcp_plain': -0.0047818333841860294, 'zcp_snip': 21.47415542602539, 'lat_1080ti_1': 0.08916890048438557, 'lat_1080ti_32': 0.057214239298789805, 'lat_1080ti_64': 0.13409705408906566, 'lat_2080ti_1': 0.07966095045680757, 'lat_2080ti_32': 0.0752030370212192, 'lat_2080ti_64': 0.11893155532229097, 'lat_essential_ph_1': 0.1320754716981132, 'lat_eyeriss': 0.33310105807492696, 'lat_fpga': 0.4731988244490841, 'lat_gold_6226': 0.3807933244085762, 'lat_gold_6240': 0.3311417934197047, 'lat_pixel2': 0.3695652173913043, 'lat_pixel3': 0.3169386871021953, 'lat_raspi4': 0.39843929658706984, 'lat_samsung_a50': 0.14736842105263157, 'lat_samsung_s7': 0.12598425196850394, 'lat_silver_4114': 0.32583290249105945, 'lat_silver_4210r': 0.33654827952127697, 'lat_titan_rtx_1': 0.06113290832157031, 'lat_titan_rtx_32': 0.06133754421451823, 'lat_titan_rtx_64': 0.06903002165612639, 'lat_titanx_1': 0.03390361349160254, 'lat_titanx_32': 0.03266740351640568, 'lat_titanx_64': 0.12971067555841426, 'lat_titanxp_1': 0.08589123531661964, 'lat_titanxp_32': 0.037175009318892276, 'lat_titanxp_64': 0.09298279878540532}
| |
FBNet_3088
|
FBNet
|
3088
|
3088
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_569[FLOAT, 16x3x3x3]
%onnx::Conv_570[FLOAT, 16]
%onnx::Conv_572[FLOAT, 16x16x1x1]
%onnx::Conv_575[FLOAT, 16x1x5x5]
%onnx::Conv_578[FLOAT, 16x16x1x1]
%onnx::Conv_581[FLOAT, 16x8x1x1]
%onnx::Conv_584[FLOAT, 16x1x3x3]
%onnx::Conv_587[FLOAT, 24x8x1x1]
%onnx::Conv_588[FLOAT, 24]
%onnx::Conv_590[FLOAT, 72x24x1x1]
%onnx::Conv_591[FLOAT, 72]
%onnx::Conv_593[FLOAT, 72x1x3x3]
%onnx::Conv_596[FLOAT, 24x72x1x1]
%onnx::Conv_599[FLOAT, 72x24x1x1]
%onnx::Conv_602[FLOAT, 72x1x3x3]
%onnx::Conv_605[FLOAT, 32x72x1x1]
%onnx::Conv_606[FLOAT, 32]
%onnx::Conv_608[FLOAT, 192x32x1x1]
%onnx::Conv_609[FLOAT, 192]
%onnx::Conv_611[FLOAT, 192x1x5x5]
%onnx::Conv_614[FLOAT, 32x192x1x1]
%onnx::Conv_617[FLOAT, 96x32x1x1]
%onnx::Conv_618[FLOAT, 96]
%onnx::Conv_620[FLOAT, 96x1x5x5]
%onnx::Conv_623[FLOAT, 32x96x1x1]
%onnx::Conv_626[FLOAT, 192x32x1x1]
%onnx::Conv_629[FLOAT, 192x1x5x5]
%onnx::Conv_632[FLOAT, 64x192x1x1]
%onnx::Conv_633[FLOAT, 64]
%onnx::Conv_635[FLOAT, 384x64x1x1]
%onnx::Conv_636[FLOAT, 384]
%onnx::Conv_638[FLOAT, 384x1x5x5]
%onnx::Conv_641[FLOAT, 64x384x1x1]
%onnx::Conv_644[FLOAT, 384x64x1x1]
%onnx::Conv_647[FLOAT, 384x1x5x5]
%onnx::Conv_650[FLOAT, 64x384x1x1]
%onnx::Conv_653[FLOAT, 64x32x1x1]
%onnx::Conv_656[FLOAT, 64x1x3x3]
%onnx::Conv_659[FLOAT, 64x32x1x1]
%onnx::Conv_662[FLOAT, 64x64x1x1]
%onnx::Conv_665[FLOAT, 64x1x3x3]
%onnx::Conv_668[FLOAT, 112x64x1x1]
%onnx::Conv_669[FLOAT, 112]
%onnx::Conv_671[FLOAT, 112x112x1x1]
%onnx::Conv_674[FLOAT, 112x1x3x3]
%onnx::Conv_677[FLOAT, 112x112x1x1]
%onnx::Conv_680[FLOAT, 336x112x1x1]
%onnx::Conv_681[FLOAT, 336]
%onnx::Conv_683[FLOAT, 336x1x3x3]
%onnx::Conv_686[FLOAT, 112x336x1x1]
%onnx::Conv_689[FLOAT, 184x112x1x1]
%onnx::Conv_690[FLOAT, 184]
%onnx::Conv_692[FLOAT, 184x184x1x1]
%onnx::Conv_695[FLOAT, 184x1x3x3]
%onnx::Conv_698[FLOAT, 184x184x1x1]
%onnx::Conv_701[FLOAT, 184x184x1x1]
%onnx::Conv_704[FLOAT, 184x1x3x3]
%onnx::Conv_707[FLOAT, 184x184x1x1]
%onnx::Conv_710[FLOAT, 184x92x1x1]
%onnx::Conv_713[FLOAT, 184x1x3x3]
%onnx::Conv_716[FLOAT, 184x92x1x1]
%onnx::Conv_719[FLOAT, 184x92x1x1]
%onnx::Conv_722[FLOAT, 184x1x5x5]
%onnx::Conv_725[FLOAT, 352x92x1x1]
%onnx::Conv_726[FLOAT, 352]
%onnx::Conv_728[FLOAT, 1504x352x1x1]
%onnx::Conv_729[FLOAT, 1504]
) {
%onnx::Conv_723 = Identity(%onnx::Conv_690)
%onnx::Conv_720 = Identity(%onnx::Conv_690)
%onnx::Conv_717 = Identity(%onnx::Conv_690)
%onnx::Conv_714 = Identity(%onnx::Conv_690)
%onnx::Conv_711 = Identity(%onnx::Conv_690)
%onnx::Conv_708 = Identity(%onnx::Conv_690)
%onnx::Conv_705 = Identity(%onnx::Conv_690)
%onnx::Conv_702 = Identity(%onnx::Conv_690)
%onnx::Conv_699 = Identity(%onnx::Conv_690)
%onnx::Conv_696 = Identity(%onnx::Conv_690)
%onnx::Conv_693 = Identity(%onnx::Conv_690)
%onnx::Conv_687 = Identity(%onnx::Conv_669)
%onnx::Conv_684 = Identity(%onnx::Conv_681)
%onnx::Conv_678 = Identity(%onnx::Conv_669)
%onnx::Conv_675 = Identity(%onnx::Conv_669)
%onnx::Conv_672 = Identity(%onnx::Conv_669)
%onnx::Conv_666 = Identity(%onnx::Conv_633)
%onnx::Conv_663 = Identity(%onnx::Conv_633)
%onnx::Conv_660 = Identity(%onnx::Conv_633)
%onnx::Conv_657 = Identity(%onnx::Conv_633)
%onnx::Conv_654 = Identity(%onnx::Conv_633)
%onnx::Conv_651 = Identity(%onnx::Conv_633)
%onnx::Conv_648 = Identity(%onnx::Conv_636)
%onnx::Conv_645 = Identity(%onnx::Conv_636)
%onnx::Conv_642 = Identity(%onnx::Conv_633)
%onnx::Conv_639 = Identity(%onnx::Conv_636)
%onnx::Conv_630 = Identity(%onnx::Conv_609)
%onnx::Conv_627 = Identity(%onnx::Conv_609)
%onnx::Conv_624 = Identity(%onnx::Conv_606)
%onnx::Conv_621 = Identity(%onnx::Conv_618)
%onnx::Conv_615 = Identity(%onnx::Conv_606)
%onnx::Conv_612 = Identity(%onnx::Conv_609)
%onnx::Conv_603 = Identity(%onnx::Conv_591)
%onnx::Conv_600 = Identity(%onnx::Conv_591)
%onnx::Conv_597 = Identity(%onnx::Conv_588)
%onnx::Conv_594 = Identity(%onnx::Conv_591)
%onnx::Conv_585 = Identity(%onnx::Conv_570)
%onnx::Conv_582 = Identity(%onnx::Conv_570)
%onnx::Conv_579 = Identity(%onnx::Conv_570)
%onnx::Conv_576 = Identity(%onnx::Conv_570)
%onnx::Conv_573 = Identity(%onnx::Conv_570)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_569, %onnx::Conv_570)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_572, %onnx::Conv_573)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_575, %onnx::Conv_576)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_578, %onnx::Conv_579)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_581, %onnx::Conv_582)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_584, %onnx::Conv_585)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_587, %onnx::Conv_588)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_590, %onnx::Conv_591)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_593, %onnx::Conv_594)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_596, %onnx::Conv_597)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_599, %onnx::Conv_600)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_602, %onnx::Conv_603)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_605, %onnx::Conv_606)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_608, %onnx::Conv_609)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_611, %onnx::Conv_612)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_614, %onnx::Conv_615)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_617, %onnx::Conv_618)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_620, %onnx::Conv_621)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_623, %onnx::Conv_624)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%567 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %567
}
|
val_accuracy
| 0
| 47,833,344
| 1,244,196
|
{'zcp_synflow': 68.28553501778259, 'zcp_zen': 59.0031852722168, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.07661762088537216, 'zcp_flops': 47833344.0, 'zcp_grad_norm': 19.42369270324707, 'zcp_grasp': -0.052025794982910156, 'zcp_jacov': -16.0422985895564, 'zcp_l2_norm': 520.5864868164062, 'zcp_nwot': 207.04347953210163, 'zcp_params': 1244196.0, 'zcp_plain': -0.005072824191302061, 'zcp_snip': 32.30129623413086, 'lat_1080ti_1': 0.3853260448850596, 'lat_1080ti_32': 0.2748436003810918, 'lat_1080ti_64': 0.15185016958418027, 'lat_2080ti_1': 0.3390687695113788, 'lat_2080ti_32': 0.23605670582532287, 'lat_2080ti_64': 0.172884455522788, 'lat_essential_ph_1': 0.09433962264150944, 'lat_eyeriss': 0.20438083362367732, 'lat_fpga': 0.18018187395481844, 'lat_gold_6226': 0.15528864152523664, 'lat_gold_6240': 0.14096664981805046, 'lat_pixel2': 0.10869565217391304, 'lat_pixel3': 0.18920708738854228, 'lat_raspi4': 0.15704613487880792, 'lat_samsung_a50': 0.07368421052631578, 'lat_samsung_s7': 0.031496062992125984, 'lat_silver_4114': 0.19322190036515247, 'lat_silver_4210r': 0.09576203098763973, 'lat_titan_rtx_1': 0.32781044491225086, 'lat_titan_rtx_32': 0.22708857367809052, 'lat_titan_rtx_64': 0.16786342420929484, 'lat_titanx_1': 0.15572459442556996, 'lat_titanx_32': 0.17871919239883066, 'lat_titanx_64': 0.1597291637895969, 'lat_titanxp_1': 0.30629248128782327, 'lat_titanxp_32': 0.21052928114365857, 'lat_titanxp_64': 0.16303502830576708}
| |
FBNet_3073
|
FBNet
|
3073
|
3073
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_705[FLOAT, 16x3x3x3]
%onnx::Conv_706[FLOAT, 16]
%onnx::Conv_708[FLOAT, 96x16x1x1]
%onnx::Conv_709[FLOAT, 96]
%onnx::Conv_711[FLOAT, 96x1x3x3]
%onnx::Conv_714[FLOAT, 16x96x1x1]
%onnx::Conv_717[FLOAT, 16x8x1x1]
%onnx::Conv_720[FLOAT, 16x1x5x5]
%onnx::Conv_723[FLOAT, 24x8x1x1]
%onnx::Conv_724[FLOAT, 24]
%onnx::Conv_726[FLOAT, 72x24x1x1]
%onnx::Conv_727[FLOAT, 72]
%onnx::Conv_729[FLOAT, 72x1x5x5]
%onnx::Conv_732[FLOAT, 24x72x1x1]
%onnx::Conv_735[FLOAT, 24x12x1x1]
%onnx::Conv_738[FLOAT, 24x1x5x5]
%onnx::Conv_741[FLOAT, 24x12x1x1]
%onnx::Conv_744[FLOAT, 72x24x1x1]
%onnx::Conv_747[FLOAT, 72x1x5x5]
%onnx::Conv_750[FLOAT, 24x72x1x1]
%onnx::Conv_753[FLOAT, 144x24x1x1]
%onnx::Conv_754[FLOAT, 144]
%onnx::Conv_756[FLOAT, 144x1x5x5]
%onnx::Conv_759[FLOAT, 32x144x1x1]
%onnx::Conv_760[FLOAT, 32]
%onnx::Conv_762[FLOAT, 32x16x1x1]
%onnx::Conv_765[FLOAT, 32x1x5x5]
%onnx::Conv_768[FLOAT, 32x16x1x1]
%onnx::Conv_771[FLOAT, 192x32x1x1]
%onnx::Conv_772[FLOAT, 192]
%onnx::Conv_774[FLOAT, 192x1x3x3]
%onnx::Conv_777[FLOAT, 32x192x1x1]
%onnx::Conv_780[FLOAT, 192x32x1x1]
%onnx::Conv_783[FLOAT, 192x1x5x5]
%onnx::Conv_786[FLOAT, 32x192x1x1]
%onnx::Conv_789[FLOAT, 32x32x1x1]
%onnx::Conv_792[FLOAT, 32x1x5x5]
%onnx::Conv_795[FLOAT, 64x32x1x1]
%onnx::Conv_796[FLOAT, 64]
%onnx::Conv_798[FLOAT, 64x64x1x1]
%onnx::Conv_801[FLOAT, 64x1x3x3]
%onnx::Conv_804[FLOAT, 64x64x1x1]
%onnx::Conv_807[FLOAT, 192x64x1x1]
%onnx::Conv_810[FLOAT, 192x1x3x3]
%onnx::Conv_813[FLOAT, 64x192x1x1]
%onnx::Conv_816[FLOAT, 64x64x1x1]
%onnx::Conv_819[FLOAT, 64x1x3x3]
%onnx::Conv_822[FLOAT, 64x64x1x1]
%onnx::Conv_825[FLOAT, 384x64x1x1]
%onnx::Conv_826[FLOAT, 384]
%onnx::Conv_828[FLOAT, 384x1x3x3]
%onnx::Conv_831[FLOAT, 112x384x1x1]
%onnx::Conv_832[FLOAT, 112]
%onnx::Conv_834[FLOAT, 336x112x1x1]
%onnx::Conv_835[FLOAT, 336]
%onnx::Conv_837[FLOAT, 336x1x3x3]
%onnx::Conv_840[FLOAT, 112x336x1x1]
%onnx::Conv_843[FLOAT, 672x112x1x1]
%onnx::Conv_844[FLOAT, 672]
%onnx::Conv_846[FLOAT, 672x1x5x5]
%onnx::Conv_849[FLOAT, 112x672x1x1]
%onnx::Conv_852[FLOAT, 112x56x1x1]
%onnx::Conv_855[FLOAT, 112x1x3x3]
%onnx::Conv_858[FLOAT, 184x56x1x1]
%onnx::Conv_859[FLOAT, 184]
%onnx::Conv_861[FLOAT, 184x92x1x1]
%onnx::Conv_864[FLOAT, 184x1x3x3]
%onnx::Conv_867[FLOAT, 184x92x1x1]
%onnx::Conv_870[FLOAT, 184x92x1x1]
%onnx::Conv_873[FLOAT, 184x1x5x5]
%onnx::Conv_876[FLOAT, 184x92x1x1]
%onnx::Conv_879[FLOAT, 184x184x1x1]
%onnx::Conv_882[FLOAT, 184x1x5x5]
%onnx::Conv_885[FLOAT, 184x184x1x1]
%onnx::Conv_888[FLOAT, 552x184x1x1]
%onnx::Conv_889[FLOAT, 552]
%onnx::Conv_891[FLOAT, 552x1x3x3]
%onnx::Conv_894[FLOAT, 352x552x1x1]
%onnx::Conv_895[FLOAT, 352]
%onnx::Conv_897[FLOAT, 1504x352x1x1]
%onnx::Conv_898[FLOAT, 1504]
) {
%onnx::Conv_892 = Identity(%onnx::Conv_889)
%onnx::Conv_886 = Identity(%onnx::Conv_859)
%onnx::Conv_883 = Identity(%onnx::Conv_859)
%onnx::Conv_880 = Identity(%onnx::Conv_859)
%onnx::Conv_877 = Identity(%onnx::Conv_859)
%onnx::Conv_874 = Identity(%onnx::Conv_859)
%onnx::Conv_871 = Identity(%onnx::Conv_859)
%onnx::Conv_868 = Identity(%onnx::Conv_859)
%onnx::Conv_865 = Identity(%onnx::Conv_859)
%onnx::Conv_862 = Identity(%onnx::Conv_859)
%onnx::Conv_856 = Identity(%onnx::Conv_832)
%onnx::Conv_853 = Identity(%onnx::Conv_832)
%onnx::Conv_850 = Identity(%onnx::Conv_832)
%onnx::Conv_847 = Identity(%onnx::Conv_844)
%onnx::Conv_841 = Identity(%onnx::Conv_832)
%onnx::Conv_838 = Identity(%onnx::Conv_835)
%onnx::Conv_829 = Identity(%onnx::Conv_826)
%onnx::Conv_823 = Identity(%onnx::Conv_796)
%onnx::Conv_820 = Identity(%onnx::Conv_796)
%onnx::Conv_817 = Identity(%onnx::Conv_796)
%onnx::Conv_814 = Identity(%onnx::Conv_796)
%onnx::Conv_811 = Identity(%onnx::Conv_772)
%onnx::Conv_808 = Identity(%onnx::Conv_772)
%onnx::Conv_805 = Identity(%onnx::Conv_796)
%onnx::Conv_802 = Identity(%onnx::Conv_796)
%onnx::Conv_799 = Identity(%onnx::Conv_796)
%onnx::Conv_793 = Identity(%onnx::Conv_760)
%onnx::Conv_790 = Identity(%onnx::Conv_760)
%onnx::Conv_787 = Identity(%onnx::Conv_760)
%onnx::Conv_784 = Identity(%onnx::Conv_772)
%onnx::Conv_781 = Identity(%onnx::Conv_772)
%onnx::Conv_778 = Identity(%onnx::Conv_760)
%onnx::Conv_775 = Identity(%onnx::Conv_772)
%onnx::Conv_769 = Identity(%onnx::Conv_760)
%onnx::Conv_766 = Identity(%onnx::Conv_760)
%onnx::Conv_763 = Identity(%onnx::Conv_760)
%onnx::Conv_757 = Identity(%onnx::Conv_754)
%onnx::Conv_751 = Identity(%onnx::Conv_724)
%onnx::Conv_748 = Identity(%onnx::Conv_727)
%onnx::Conv_745 = Identity(%onnx::Conv_727)
%onnx::Conv_742 = Identity(%onnx::Conv_724)
%onnx::Conv_739 = Identity(%onnx::Conv_724)
%onnx::Conv_736 = Identity(%onnx::Conv_724)
%onnx::Conv_733 = Identity(%onnx::Conv_724)
%onnx::Conv_730 = Identity(%onnx::Conv_727)
%onnx::Conv_721 = Identity(%onnx::Conv_706)
%onnx::Conv_718 = Identity(%onnx::Conv_706)
%onnx::Conv_715 = Identity(%onnx::Conv_706)
%onnx::Conv_712 = Identity(%onnx::Conv_709)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_705, %onnx::Conv_706)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_858, %onnx::Conv_859)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_861, %onnx::Conv_862)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_output_0 = Reshape(%/cells.18/nl/Relu_output_0, %/cells.18/shuffle/Constant_output_0)
%/cells.18/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.18/shuffle/Reshape_output_0)
%/cells.18/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_1_output_0 = Reshape(%/cells.18/shuffle/Transpose_output_0, %/cells.18/shuffle/Constant_1_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/shuffle/Reshape_1_output_0, %onnx::Conv_864, %onnx::Conv_865)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_867, %onnx::Conv_868)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_870, %onnx::Conv_871)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_873, %onnx::Conv_874)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_876, %onnx::Conv_877)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_879, %onnx::Conv_880)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_882, %onnx::Conv_883)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_885, %onnx::Conv_886)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_888, %onnx::Conv_889)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_891, %onnx::Conv_892)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_894, %onnx::Conv_895)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_897, %onnx::Conv_898)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%703 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %703
}
|
val_accuracy
| 0
| 74,722,688
| 1,592,428
|
{'zcp_synflow': 78.20240363195846, 'zcp_zen': 69.72293853759766, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.16513793170452118, 'zcp_flops': 74722688.0, 'zcp_grad_norm': 27.87158966064453, 'zcp_grasp': 0.5196990966796875, 'zcp_jacov': -16.057427271902426, 'zcp_l2_norm': 615.664306640625, 'zcp_nwot': 215.67067676189492, 'zcp_params': 1592428.0, 'zcp_plain': -0.00370492204092443, 'zcp_snip': 50.38851547241211, 'lat_1080ti_1': 0.6633275019638941, 'lat_1080ti_32': 0.7197289086555679, 'lat_1080ti_64': 0.6639996732863623, 'lat_2080ti_1': 0.7545876131184327, 'lat_2080ti_32': 0.6686468830021325, 'lat_2080ti_64': 0.6587905401922374, 'lat_essential_ph_1': 0.33962264150943394, 'lat_eyeriss': 0.5207397355729557, 'lat_fpga': 0.5147587053107591, 'lat_gold_6226': 0.3225016821122442, 'lat_gold_6240': 0.48753598546455995, 'lat_pixel2': 0.34782608695652173, 'lat_pixel3': 0.5646854301488122, 'lat_raspi4': 0.5714015495590256, 'lat_samsung_a50': 0.2, 'lat_samsung_s7': 0.12598425196850394, 'lat_silver_4114': 0.5016062647050695, 'lat_silver_4210r': 0.5306387680487192, 'lat_titan_rtx_1': 0.7078106974806684, 'lat_titan_rtx_32': 0.619083896023635, 'lat_titan_rtx_64': 0.6744300546582731, 'lat_titanx_1': 0.37391440276474286, 'lat_titanx_32': 0.6560500223825495, 'lat_titanx_64': 0.6499190225428952, 'lat_titanxp_1': 0.6568134454520368, 'lat_titanxp_32': 0.6660147417949538, 'lat_titanxp_64': 0.6683250594124496}
| |
FBNet_390
|
FBNet
|
390
|
390
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_632[FLOAT, 16x3x3x3]
%onnx::Conv_633[FLOAT, 16]
%onnx::Conv_635[FLOAT, 48x16x1x1]
%onnx::Conv_636[FLOAT, 48]
%onnx::Conv_638[FLOAT, 48x1x5x5]
%onnx::Conv_641[FLOAT, 16x48x1x1]
%onnx::Conv_644[FLOAT, 96x16x1x1]
%onnx::Conv_645[FLOAT, 96]
%onnx::Conv_647[FLOAT, 96x1x5x5]
%onnx::Conv_650[FLOAT, 24x96x1x1]
%onnx::Conv_651[FLOAT, 24]
%onnx::Conv_653[FLOAT, 72x24x1x1]
%onnx::Conv_654[FLOAT, 72]
%onnx::Conv_656[FLOAT, 72x1x5x5]
%onnx::Conv_659[FLOAT, 24x72x1x1]
%onnx::Conv_662[FLOAT, 144x24x1x1]
%onnx::Conv_663[FLOAT, 144]
%onnx::Conv_665[FLOAT, 144x1x5x5]
%onnx::Conv_668[FLOAT, 24x144x1x1]
%onnx::Conv_671[FLOAT, 72x24x1x1]
%onnx::Conv_674[FLOAT, 72x1x5x5]
%onnx::Conv_677[FLOAT, 24x72x1x1]
%onnx::Conv_680[FLOAT, 144x24x1x1]
%onnx::Conv_683[FLOAT, 144x1x5x5]
%onnx::Conv_686[FLOAT, 32x144x1x1]
%onnx::Conv_687[FLOAT, 32]
%onnx::Conv_689[FLOAT, 32x32x1x1]
%onnx::Conv_692[FLOAT, 32x1x3x3]
%onnx::Conv_695[FLOAT, 32x32x1x1]
%onnx::Conv_698[FLOAT, 32x32x1x1]
%onnx::Conv_701[FLOAT, 32x1x3x3]
%onnx::Conv_704[FLOAT, 32x32x1x1]
%onnx::Conv_707[FLOAT, 96x32x1x1]
%onnx::Conv_710[FLOAT, 96x1x5x5]
%onnx::Conv_713[FLOAT, 32x96x1x1]
%onnx::Conv_716[FLOAT, 64x32x1x1]
%onnx::Conv_717[FLOAT, 64]
%onnx::Conv_719[FLOAT, 384x64x1x1]
%onnx::Conv_720[FLOAT, 384]
%onnx::Conv_722[FLOAT, 384x1x5x5]
%onnx::Conv_725[FLOAT, 64x384x1x1]
%onnx::Conv_728[FLOAT, 384x64x1x1]
%onnx::Conv_731[FLOAT, 384x1x3x3]
%onnx::Conv_734[FLOAT, 64x384x1x1]
%onnx::Conv_737[FLOAT, 64x64x1x1]
%onnx::Conv_740[FLOAT, 64x1x5x5]
%onnx::Conv_743[FLOAT, 64x64x1x1]
%onnx::Conv_746[FLOAT, 64x32x1x1]
%onnx::Conv_749[FLOAT, 64x1x3x3]
%onnx::Conv_752[FLOAT, 112x32x1x1]
%onnx::Conv_753[FLOAT, 112]
%onnx::Conv_755[FLOAT, 336x112x1x1]
%onnx::Conv_756[FLOAT, 336]
%onnx::Conv_758[FLOAT, 336x1x3x3]
%onnx::Conv_761[FLOAT, 112x336x1x1]
%onnx::Conv_764[FLOAT, 112x56x1x1]
%onnx::Conv_767[FLOAT, 112x1x3x3]
%onnx::Conv_770[FLOAT, 112x56x1x1]
%onnx::Conv_773[FLOAT, 672x112x1x1]
%onnx::Conv_774[FLOAT, 672]
%onnx::Conv_776[FLOAT, 672x1x5x5]
%onnx::Conv_779[FLOAT, 184x672x1x1]
%onnx::Conv_780[FLOAT, 184]
%onnx::Conv_782[FLOAT, 1104x184x1x1]
%onnx::Conv_783[FLOAT, 1104]
%onnx::Conv_785[FLOAT, 1104x1x5x5]
%onnx::Conv_788[FLOAT, 184x1104x1x1]
%onnx::Conv_791[FLOAT, 552x184x1x1]
%onnx::Conv_792[FLOAT, 552]
%onnx::Conv_794[FLOAT, 552x1x3x3]
%onnx::Conv_797[FLOAT, 184x552x1x1]
%onnx::Conv_800[FLOAT, 552x184x1x1]
%onnx::Conv_803[FLOAT, 552x1x3x3]
%onnx::Conv_806[FLOAT, 184x552x1x1]
%onnx::Conv_809[FLOAT, 184x92x1x1]
%onnx::Conv_812[FLOAT, 184x1x5x5]
%onnx::Conv_815[FLOAT, 352x92x1x1]
%onnx::Conv_816[FLOAT, 352]
%onnx::Conv_818[FLOAT, 1504x352x1x1]
%onnx::Conv_819[FLOAT, 1504]
) {
%onnx::Conv_813 = Identity(%onnx::Conv_780)
%onnx::Conv_810 = Identity(%onnx::Conv_780)
%onnx::Conv_807 = Identity(%onnx::Conv_780)
%onnx::Conv_804 = Identity(%onnx::Conv_792)
%onnx::Conv_801 = Identity(%onnx::Conv_792)
%onnx::Conv_798 = Identity(%onnx::Conv_780)
%onnx::Conv_795 = Identity(%onnx::Conv_792)
%onnx::Conv_789 = Identity(%onnx::Conv_780)
%onnx::Conv_786 = Identity(%onnx::Conv_783)
%onnx::Conv_777 = Identity(%onnx::Conv_774)
%onnx::Conv_771 = Identity(%onnx::Conv_753)
%onnx::Conv_768 = Identity(%onnx::Conv_753)
%onnx::Conv_765 = Identity(%onnx::Conv_753)
%onnx::Conv_762 = Identity(%onnx::Conv_753)
%onnx::Conv_759 = Identity(%onnx::Conv_756)
%onnx::Conv_750 = Identity(%onnx::Conv_717)
%onnx::Conv_747 = Identity(%onnx::Conv_717)
%onnx::Conv_744 = Identity(%onnx::Conv_717)
%onnx::Conv_741 = Identity(%onnx::Conv_717)
%onnx::Conv_738 = Identity(%onnx::Conv_717)
%onnx::Conv_735 = Identity(%onnx::Conv_717)
%onnx::Conv_732 = Identity(%onnx::Conv_720)
%onnx::Conv_729 = Identity(%onnx::Conv_720)
%onnx::Conv_726 = Identity(%onnx::Conv_717)
%onnx::Conv_723 = Identity(%onnx::Conv_720)
%onnx::Conv_714 = Identity(%onnx::Conv_687)
%onnx::Conv_711 = Identity(%onnx::Conv_645)
%onnx::Conv_708 = Identity(%onnx::Conv_645)
%onnx::Conv_705 = Identity(%onnx::Conv_687)
%onnx::Conv_702 = Identity(%onnx::Conv_687)
%onnx::Conv_699 = Identity(%onnx::Conv_687)
%onnx::Conv_696 = Identity(%onnx::Conv_687)
%onnx::Conv_693 = Identity(%onnx::Conv_687)
%onnx::Conv_690 = Identity(%onnx::Conv_687)
%onnx::Conv_684 = Identity(%onnx::Conv_663)
%onnx::Conv_681 = Identity(%onnx::Conv_663)
%onnx::Conv_678 = Identity(%onnx::Conv_651)
%onnx::Conv_675 = Identity(%onnx::Conv_654)
%onnx::Conv_672 = Identity(%onnx::Conv_654)
%onnx::Conv_669 = Identity(%onnx::Conv_651)
%onnx::Conv_666 = Identity(%onnx::Conv_663)
%onnx::Conv_660 = Identity(%onnx::Conv_651)
%onnx::Conv_657 = Identity(%onnx::Conv_654)
%onnx::Conv_648 = Identity(%onnx::Conv_645)
%onnx::Conv_642 = Identity(%onnx::Conv_633)
%onnx::Conv_639 = Identity(%onnx::Conv_636)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_632, %onnx::Conv_633)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.8/Add_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.9/relu/Relu_output_0 = Relu(%/cells.9/conv/Conv_output_0)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/relu/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/relu/Relu_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_803, %onnx::Conv_804)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_806, %onnx::Conv_807)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_809, %onnx::Conv_810)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_812, %onnx::Conv_813)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_815, %onnx::Conv_816)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_818, %onnx::Conv_819)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%630 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %630
}
|
val_accuracy
| 0
| 88,611,456
| 2,103,628
|
{'zcp_synflow': 80.84977819501682, 'zcp_zen': 71.4151840209961, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.1314721703529358, 'zcp_flops': 88611456.0, 'zcp_grad_norm': 33.89366912841797, 'zcp_grasp': 0.055789947509765625, 'zcp_jacov': -16.05014106150748, 'zcp_l2_norm': 667.826171875, 'zcp_nwot': 218.06393981605845, 'zcp_params': 2103628.0, 'zcp_plain': 0.0017923059640452266, 'zcp_snip': 56.4675407409668, 'lat_1080ti_1': 0.6064394075642389, 'lat_1080ti_32': 0.7795403820552411, 'lat_1080ti_64': 0.779621353880823, 'lat_2080ti_1': 0.5949031222652911, 'lat_2080ti_32': 0.6843885723732912, 'lat_2080ti_64': 0.7429066505581722, 'lat_essential_ph_1': 0.5094339622641509, 'lat_eyeriss': 0.7438615150459353, 'lat_fpga': 0.6171741260056006, 'lat_gold_6226': 0.4872679080250786, 'lat_gold_6240': 0.5677695500464569, 'lat_pixel2': 0.5217391304347826, 'lat_pixel3': 0.767924027041577, 'lat_raspi4': 0.7279437550247941, 'lat_samsung_a50': 0.29473684210526313, 'lat_samsung_s7': 0.2992125984251969, 'lat_silver_4114': 0.5746686661965192, 'lat_silver_4210r': 0.6041794306438198, 'lat_titan_rtx_1': 0.5914534854835718, 'lat_titan_rtx_32': 0.723272055516557, 'lat_titan_rtx_64': 0.7634046616180876, 'lat_titanx_1': 0.3151025079381866, 'lat_titanx_32': 0.7422282274331932, 'lat_titanx_64': 0.7455177304263685, 'lat_titanxp_1': 0.5461021595181481, 'lat_titanxp_32': 0.7456780726893925, 'lat_titanxp_64': 0.8055246630731605}
| |
FBNet_603
|
FBNet
|
603
|
603
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_572[FLOAT, 16x3x3x3]
%onnx::Conv_573[FLOAT, 16]
%onnx::Conv_575[FLOAT, 16x8x1x1]
%onnx::Conv_578[FLOAT, 16x1x3x3]
%onnx::Conv_581[FLOAT, 16x8x1x1]
%onnx::Conv_584[FLOAT, 16x16x1x1]
%onnx::Conv_587[FLOAT, 16x1x5x5]
%onnx::Conv_590[FLOAT, 24x16x1x1]
%onnx::Conv_591[FLOAT, 24]
%onnx::Conv_593[FLOAT, 24x12x1x1]
%onnx::Conv_596[FLOAT, 24x1x5x5]
%onnx::Conv_599[FLOAT, 24x12x1x1]
%onnx::Conv_602[FLOAT, 144x24x1x1]
%onnx::Conv_603[FLOAT, 144]
%onnx::Conv_605[FLOAT, 144x1x5x5]
%onnx::Conv_608[FLOAT, 24x144x1x1]
%onnx::Conv_611[FLOAT, 144x24x1x1]
%onnx::Conv_614[FLOAT, 144x1x3x3]
%onnx::Conv_617[FLOAT, 24x144x1x1]
%onnx::Conv_620[FLOAT, 32x24x1x1]
%onnx::Conv_621[FLOAT, 32]
%onnx::Conv_623[FLOAT, 32x16x1x1]
%onnx::Conv_626[FLOAT, 32x1x5x5]
%onnx::Conv_629[FLOAT, 64x16x1x1]
%onnx::Conv_630[FLOAT, 64]
%onnx::Conv_632[FLOAT, 192x64x1x1]
%onnx::Conv_633[FLOAT, 192]
%onnx::Conv_635[FLOAT, 192x1x5x5]
%onnx::Conv_638[FLOAT, 64x192x1x1]
%onnx::Conv_641[FLOAT, 64x32x1x1]
%onnx::Conv_644[FLOAT, 64x1x5x5]
%onnx::Conv_647[FLOAT, 64x32x1x1]
%onnx::Conv_650[FLOAT, 192x64x1x1]
%onnx::Conv_653[FLOAT, 192x1x3x3]
%onnx::Conv_656[FLOAT, 64x192x1x1]
%onnx::Conv_659[FLOAT, 64x64x1x1]
%onnx::Conv_662[FLOAT, 64x1x5x5]
%onnx::Conv_665[FLOAT, 112x64x1x1]
%onnx::Conv_666[FLOAT, 112]
%onnx::Conv_668[FLOAT, 336x112x1x1]
%onnx::Conv_669[FLOAT, 336]
%onnx::Conv_671[FLOAT, 336x1x3x3]
%onnx::Conv_674[FLOAT, 112x336x1x1]
%onnx::Conv_677[FLOAT, 336x112x1x1]
%onnx::Conv_680[FLOAT, 336x1x3x3]
%onnx::Conv_683[FLOAT, 112x336x1x1]
%onnx::Conv_686[FLOAT, 184x112x1x1]
%onnx::Conv_687[FLOAT, 184]
%onnx::Conv_689[FLOAT, 184x184x1x1]
%onnx::Conv_692[FLOAT, 184x1x5x5]
%onnx::Conv_695[FLOAT, 184x184x1x1]
%onnx::Conv_698[FLOAT, 1104x184x1x1]
%onnx::Conv_699[FLOAT, 1104]
%onnx::Conv_701[FLOAT, 1104x1x5x5]
%onnx::Conv_704[FLOAT, 184x1104x1x1]
%onnx::Conv_707[FLOAT, 184x92x1x1]
%onnx::Conv_710[FLOAT, 184x1x5x5]
%onnx::Conv_713[FLOAT, 184x92x1x1]
%onnx::Conv_716[FLOAT, 184x184x1x1]
%onnx::Conv_719[FLOAT, 184x1x5x5]
%onnx::Conv_722[FLOAT, 352x184x1x1]
%onnx::Conv_723[FLOAT, 352]
%onnx::Conv_725[FLOAT, 1504x352x1x1]
%onnx::Conv_726[FLOAT, 1504]
) {
%onnx::Conv_720 = Identity(%onnx::Conv_687)
%onnx::Conv_717 = Identity(%onnx::Conv_687)
%onnx::Conv_714 = Identity(%onnx::Conv_687)
%onnx::Conv_711 = Identity(%onnx::Conv_687)
%onnx::Conv_708 = Identity(%onnx::Conv_687)
%onnx::Conv_705 = Identity(%onnx::Conv_687)
%onnx::Conv_702 = Identity(%onnx::Conv_699)
%onnx::Conv_696 = Identity(%onnx::Conv_687)
%onnx::Conv_693 = Identity(%onnx::Conv_687)
%onnx::Conv_690 = Identity(%onnx::Conv_687)
%onnx::Conv_684 = Identity(%onnx::Conv_666)
%onnx::Conv_681 = Identity(%onnx::Conv_669)
%onnx::Conv_678 = Identity(%onnx::Conv_669)
%onnx::Conv_675 = Identity(%onnx::Conv_666)
%onnx::Conv_672 = Identity(%onnx::Conv_669)
%onnx::Conv_663 = Identity(%onnx::Conv_630)
%onnx::Conv_660 = Identity(%onnx::Conv_630)
%onnx::Conv_657 = Identity(%onnx::Conv_630)
%onnx::Conv_654 = Identity(%onnx::Conv_633)
%onnx::Conv_651 = Identity(%onnx::Conv_633)
%onnx::Conv_648 = Identity(%onnx::Conv_630)
%onnx::Conv_645 = Identity(%onnx::Conv_630)
%onnx::Conv_642 = Identity(%onnx::Conv_630)
%onnx::Conv_639 = Identity(%onnx::Conv_630)
%onnx::Conv_636 = Identity(%onnx::Conv_633)
%onnx::Conv_627 = Identity(%onnx::Conv_621)
%onnx::Conv_624 = Identity(%onnx::Conv_621)
%onnx::Conv_618 = Identity(%onnx::Conv_591)
%onnx::Conv_615 = Identity(%onnx::Conv_603)
%onnx::Conv_612 = Identity(%onnx::Conv_603)
%onnx::Conv_609 = Identity(%onnx::Conv_591)
%onnx::Conv_606 = Identity(%onnx::Conv_603)
%onnx::Conv_600 = Identity(%onnx::Conv_591)
%onnx::Conv_597 = Identity(%onnx::Conv_591)
%onnx::Conv_594 = Identity(%onnx::Conv_591)
%onnx::Conv_588 = Identity(%onnx::Conv_573)
%onnx::Conv_585 = Identity(%onnx::Conv_573)
%onnx::Conv_582 = Identity(%onnx::Conv_573)
%onnx::Conv_579 = Identity(%onnx::Conv_573)
%onnx::Conv_576 = Identity(%onnx::Conv_573)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_572, %onnx::Conv_573)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_575, %onnx::Conv_576)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_578, %onnx::Conv_579)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_581, %onnx::Conv_582)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_584, %onnx::Conv_585)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_587, %onnx::Conv_588)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_590, %onnx::Conv_591)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_593, %onnx::Conv_594)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_596, %onnx::Conv_597)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_599, %onnx::Conv_600)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_602, %onnx::Conv_603)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_605, %onnx::Conv_606)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_608, %onnx::Conv_609)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_611, %onnx::Conv_612)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_614, %onnx::Conv_615)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_617, %onnx::Conv_618)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.4/Add_output_0, %onnx::Conv_620, %onnx::Conv_621)
%/cells.5/relu/Relu_output_0 = Relu(%/cells.5/conv/Conv_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/relu/Relu_output_0, %onnx::Conv_623, %onnx::Conv_624)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%570 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %570
}
|
val_accuracy
| 0
| 59,631,616
| 1,623,620
|
{'zcp_synflow': 66.12977094087357, 'zcp_zen': 57.54740524291992, 'zcp_epe_nas': 15.470776103064177, 'zcp_fisher': 0.029415670782327652, 'zcp_flops': 59631616.0, 'zcp_grad_norm': 13.50589370727539, 'zcp_grasp': -0.012964248657226562, 'zcp_jacov': -16.067136987297395, 'zcp_l2_norm': 514.3754272460938, 'zcp_nwot': 211.07126873231175, 'zcp_params': 1623620.0, 'zcp_plain': -0.0021178117021918297, 'zcp_snip': 22.454517364501953, 'lat_1080ti_1': 0.23857162491218722, 'lat_1080ti_32': 0.35666179618160954, 'lat_1080ti_64': 0.3920255082350506, 'lat_2080ti_1': 0.2779206532497129, 'lat_2080ti_32': 0.39291122977089626, 'lat_2080ti_64': 0.385609201731142, 'lat_essential_ph_1': 0.16981132075471697, 'lat_eyeriss': 0.34568976582986444, 'lat_fpga': 0.36621181685945825, 'lat_gold_6226': 0.21860976734482448, 'lat_gold_6240': 0.26949263184906125, 'lat_pixel2': 0.1956521739130435, 'lat_pixel3': 0.384490789911587, 'lat_raspi4': 0.41488485310266854, 'lat_samsung_a50': 0.1368421052631579, 'lat_samsung_s7': 0.6377952755905512, 'lat_silver_4114': 0.27958893988303496, 'lat_silver_4210r': 0.2666351176397183, 'lat_titan_rtx_1': 0.274517099556436, 'lat_titan_rtx_32': 0.38015312152138936, 'lat_titan_rtx_64': 0.38532092804538853, 'lat_titanx_1': 0.14454519273059005, 'lat_titanx_32': 0.3671386224440833, 'lat_titanx_64': 0.3664101940584062, 'lat_titanxp_1': 0.25788020745945195, 'lat_titanxp_32': 0.37175421563325295, 'lat_titanxp_64': 0.40400401687186566}
| |
FBNet_4841
|
FBNet
|
4841
|
4841
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_652[FLOAT, 16x3x3x3]
%onnx::Conv_653[FLOAT, 16]
%onnx::Conv_655[FLOAT, 24x16x1x1]
%onnx::Conv_656[FLOAT, 24]
%onnx::Conv_658[FLOAT, 72x24x1x1]
%onnx::Conv_659[FLOAT, 72]
%onnx::Conv_661[FLOAT, 72x1x5x5]
%onnx::Conv_664[FLOAT, 24x72x1x1]
%onnx::Conv_667[FLOAT, 24x24x1x1]
%onnx::Conv_670[FLOAT, 24x1x3x3]
%onnx::Conv_673[FLOAT, 24x24x1x1]
%onnx::Conv_676[FLOAT, 24x12x1x1]
%onnx::Conv_679[FLOAT, 24x1x3x3]
%onnx::Conv_682[FLOAT, 24x12x1x1]
%onnx::Conv_685[FLOAT, 72x24x1x1]
%onnx::Conv_688[FLOAT, 72x1x3x3]
%onnx::Conv_691[FLOAT, 32x72x1x1]
%onnx::Conv_692[FLOAT, 32]
%onnx::Conv_694[FLOAT, 32x16x1x1]
%onnx::Conv_697[FLOAT, 32x1x3x3]
%onnx::Conv_700[FLOAT, 32x16x1x1]
%onnx::Conv_703[FLOAT, 192x32x1x1]
%onnx::Conv_704[FLOAT, 192]
%onnx::Conv_706[FLOAT, 192x1x3x3]
%onnx::Conv_709[FLOAT, 32x192x1x1]
%onnx::Conv_712[FLOAT, 32x16x1x1]
%onnx::Conv_715[FLOAT, 32x1x5x5]
%onnx::Conv_718[FLOAT, 32x16x1x1]
%onnx::Conv_721[FLOAT, 96x32x1x1]
%onnx::Conv_722[FLOAT, 96]
%onnx::Conv_724[FLOAT, 96x1x3x3]
%onnx::Conv_727[FLOAT, 64x96x1x1]
%onnx::Conv_728[FLOAT, 64]
%onnx::Conv_730[FLOAT, 192x64x1x1]
%onnx::Conv_733[FLOAT, 192x1x3x3]
%onnx::Conv_736[FLOAT, 64x192x1x1]
%onnx::Conv_739[FLOAT, 192x64x1x1]
%onnx::Conv_742[FLOAT, 192x1x5x5]
%onnx::Conv_745[FLOAT, 64x192x1x1]
%onnx::Conv_748[FLOAT, 64x32x1x1]
%onnx::Conv_751[FLOAT, 64x1x5x5]
%onnx::Conv_754[FLOAT, 112x32x1x1]
%onnx::Conv_755[FLOAT, 112]
%onnx::Conv_757[FLOAT, 336x112x1x1]
%onnx::Conv_758[FLOAT, 336]
%onnx::Conv_760[FLOAT, 336x1x5x5]
%onnx::Conv_763[FLOAT, 112x336x1x1]
%onnx::Conv_766[FLOAT, 112x56x1x1]
%onnx::Conv_769[FLOAT, 112x1x5x5]
%onnx::Conv_772[FLOAT, 112x56x1x1]
%onnx::Conv_775[FLOAT, 112x56x1x1]
%onnx::Conv_778[FLOAT, 112x1x5x5]
%onnx::Conv_781[FLOAT, 184x56x1x1]
%onnx::Conv_782[FLOAT, 184]
%onnx::Conv_784[FLOAT, 184x184x1x1]
%onnx::Conv_787[FLOAT, 184x1x3x3]
%onnx::Conv_790[FLOAT, 184x184x1x1]
%onnx::Conv_793[FLOAT, 1104x184x1x1]
%onnx::Conv_794[FLOAT, 1104]
%onnx::Conv_796[FLOAT, 1104x1x3x3]
%onnx::Conv_799[FLOAT, 184x1104x1x1]
%onnx::Conv_802[FLOAT, 184x184x1x1]
%onnx::Conv_805[FLOAT, 184x1x5x5]
%onnx::Conv_808[FLOAT, 184x184x1x1]
%onnx::Conv_811[FLOAT, 184x92x1x1]
%onnx::Conv_814[FLOAT, 184x1x5x5]
%onnx::Conv_817[FLOAT, 352x92x1x1]
%onnx::Conv_818[FLOAT, 352]
%onnx::Conv_820[FLOAT, 1504x352x1x1]
%onnx::Conv_821[FLOAT, 1504]
) {
%onnx::Conv_815 = Identity(%onnx::Conv_782)
%onnx::Conv_812 = Identity(%onnx::Conv_782)
%onnx::Conv_809 = Identity(%onnx::Conv_782)
%onnx::Conv_806 = Identity(%onnx::Conv_782)
%onnx::Conv_803 = Identity(%onnx::Conv_782)
%onnx::Conv_800 = Identity(%onnx::Conv_782)
%onnx::Conv_797 = Identity(%onnx::Conv_794)
%onnx::Conv_791 = Identity(%onnx::Conv_782)
%onnx::Conv_788 = Identity(%onnx::Conv_782)
%onnx::Conv_785 = Identity(%onnx::Conv_782)
%onnx::Conv_779 = Identity(%onnx::Conv_755)
%onnx::Conv_776 = Identity(%onnx::Conv_755)
%onnx::Conv_773 = Identity(%onnx::Conv_755)
%onnx::Conv_770 = Identity(%onnx::Conv_755)
%onnx::Conv_767 = Identity(%onnx::Conv_755)
%onnx::Conv_764 = Identity(%onnx::Conv_755)
%onnx::Conv_761 = Identity(%onnx::Conv_758)
%onnx::Conv_752 = Identity(%onnx::Conv_728)
%onnx::Conv_749 = Identity(%onnx::Conv_728)
%onnx::Conv_746 = Identity(%onnx::Conv_728)
%onnx::Conv_743 = Identity(%onnx::Conv_704)
%onnx::Conv_740 = Identity(%onnx::Conv_704)
%onnx::Conv_737 = Identity(%onnx::Conv_728)
%onnx::Conv_734 = Identity(%onnx::Conv_704)
%onnx::Conv_731 = Identity(%onnx::Conv_704)
%onnx::Conv_725 = Identity(%onnx::Conv_722)
%onnx::Conv_719 = Identity(%onnx::Conv_692)
%onnx::Conv_716 = Identity(%onnx::Conv_692)
%onnx::Conv_713 = Identity(%onnx::Conv_692)
%onnx::Conv_710 = Identity(%onnx::Conv_692)
%onnx::Conv_707 = Identity(%onnx::Conv_704)
%onnx::Conv_701 = Identity(%onnx::Conv_692)
%onnx::Conv_698 = Identity(%onnx::Conv_692)
%onnx::Conv_695 = Identity(%onnx::Conv_692)
%onnx::Conv_689 = Identity(%onnx::Conv_659)
%onnx::Conv_686 = Identity(%onnx::Conv_659)
%onnx::Conv_683 = Identity(%onnx::Conv_656)
%onnx::Conv_680 = Identity(%onnx::Conv_656)
%onnx::Conv_677 = Identity(%onnx::Conv_656)
%onnx::Conv_674 = Identity(%onnx::Conv_656)
%onnx::Conv_671 = Identity(%onnx::Conv_656)
%onnx::Conv_668 = Identity(%onnx::Conv_656)
%onnx::Conv_665 = Identity(%onnx::Conv_656)
%onnx::Conv_662 = Identity(%onnx::Conv_659)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_652, %onnx::Conv_653)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.1/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_655, %onnx::Conv_656)
%/cells.1/relu/Relu_output_0 = Relu(%/cells.1/conv/Conv_output_0)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/relu/Relu_output_0, %onnx::Conv_658, %onnx::Conv_659)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/relu/Relu_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%650 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %650
}
|
val_accuracy
| 0
| 48,150,144
| 1,532,700
|
{'zcp_synflow': 67.621412230475, 'zcp_zen': 59.778507232666016, 'zcp_epe_nas': 6.711672362650362, 'zcp_fisher': 0.05986329913139343, 'zcp_flops': 48150144.0, 'zcp_grad_norm': 18.378318786621094, 'zcp_grasp': -0.07268905639648438, 'zcp_jacov': -16.05380640076861, 'zcp_l2_norm': 527.6771240234375, 'zcp_nwot': 206.73723717214557, 'zcp_params': 1532700.0, 'zcp_plain': 0.006099556107074022, 'zcp_snip': 30.792123794555664, 'lat_1080ti_1': 0.5502530126036239, 'lat_1080ti_32': 0.3880822397949532, 'lat_1080ti_64': 0.25627501101048034, 'lat_2080ti_1': 0.5073555450343917, 'lat_2080ti_32': 0.40473262330838633, 'lat_2080ti_64': 0.27923029165533225, 'lat_essential_ph_1': 0.07547169811320754, 'lat_eyeriss': 0.2000165037683603, 'lat_fpga': 0.20195133535855295, 'lat_gold_6226': 0.19894895158324785, 'lat_gold_6240': 0.31579420179944856, 'lat_pixel2': 0.2391304347826087, 'lat_pixel3': 0.20930006714746122, 'lat_raspi4': 0.22526404408375547, 'lat_samsung_a50': 0.10526315789473684, 'lat_samsung_s7': 0.09448818897637795, 'lat_silver_4114': 0.32607517602146446, 'lat_silver_4210r': 0.3372519335461382, 'lat_titan_rtx_1': 0.47024863159434077, 'lat_titan_rtx_32': 0.39523647400138284, 'lat_titan_rtx_64': 0.2897580976212994, 'lat_titanx_1': 0.2471324184357146, 'lat_titanx_32': 0.3485078328125113, 'lat_titanx_64': 0.24666627622705828, 'lat_titanxp_1': 0.43148220451682545, 'lat_titanxp_32': 0.3635881597499312, 'lat_titanxp_64': 0.2529205298378879}
| |
FBNet_4549
|
FBNet
|
4549
|
4549
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_585[FLOAT, 16x3x3x3]
%onnx::Conv_586[FLOAT, 16]
%onnx::Conv_588[FLOAT, 48x16x1x1]
%onnx::Conv_589[FLOAT, 48]
%onnx::Conv_591[FLOAT, 48x1x5x5]
%onnx::Conv_594[FLOAT, 16x48x1x1]
%onnx::Conv_597[FLOAT, 96x16x1x1]
%onnx::Conv_598[FLOAT, 96]
%onnx::Conv_600[FLOAT, 96x1x5x5]
%onnx::Conv_603[FLOAT, 24x96x1x1]
%onnx::Conv_604[FLOAT, 24]
%onnx::Conv_606[FLOAT, 72x24x1x1]
%onnx::Conv_607[FLOAT, 72]
%onnx::Conv_609[FLOAT, 72x1x5x5]
%onnx::Conv_612[FLOAT, 24x72x1x1]
%onnx::Conv_615[FLOAT, 144x24x1x1]
%onnx::Conv_616[FLOAT, 144]
%onnx::Conv_618[FLOAT, 144x1x5x5]
%onnx::Conv_621[FLOAT, 24x144x1x1]
%onnx::Conv_624[FLOAT, 144x24x1x1]
%onnx::Conv_627[FLOAT, 144x1x5x5]
%onnx::Conv_630[FLOAT, 24x144x1x1]
%onnx::Conv_633[FLOAT, 144x24x1x1]
%onnx::Conv_636[FLOAT, 144x1x5x5]
%onnx::Conv_639[FLOAT, 32x144x1x1]
%onnx::Conv_640[FLOAT, 32]
%onnx::Conv_642[FLOAT, 32x16x1x1]
%onnx::Conv_645[FLOAT, 32x1x3x3]
%onnx::Conv_648[FLOAT, 32x16x1x1]
%onnx::Conv_651[FLOAT, 96x32x1x1]
%onnx::Conv_654[FLOAT, 96x1x5x5]
%onnx::Conv_657[FLOAT, 32x96x1x1]
%onnx::Conv_660[FLOAT, 32x16x1x1]
%onnx::Conv_663[FLOAT, 32x1x5x5]
%onnx::Conv_666[FLOAT, 32x16x1x1]
%onnx::Conv_669[FLOAT, 192x32x1x1]
%onnx::Conv_670[FLOAT, 192]
%onnx::Conv_672[FLOAT, 192x1x5x5]
%onnx::Conv_675[FLOAT, 64x192x1x1]
%onnx::Conv_676[FLOAT, 64]
%onnx::Conv_678[FLOAT, 192x64x1x1]
%onnx::Conv_681[FLOAT, 192x1x5x5]
%onnx::Conv_684[FLOAT, 64x192x1x1]
%onnx::Conv_687[FLOAT, 64x32x1x1]
%onnx::Conv_690[FLOAT, 64x1x3x3]
%onnx::Conv_693[FLOAT, 64x32x1x1]
%onnx::Conv_696[FLOAT, 64x32x1x1]
%onnx::Conv_699[FLOAT, 64x1x3x3]
%onnx::Conv_702[FLOAT, 64x32x1x1]
%onnx::Conv_705[FLOAT, 64x64x1x1]
%onnx::Conv_708[FLOAT, 64x1x3x3]
%onnx::Conv_711[FLOAT, 112x64x1x1]
%onnx::Conv_712[FLOAT, 112]
%onnx::Conv_714[FLOAT, 112x112x1x1]
%onnx::Conv_717[FLOAT, 112x1x5x5]
%onnx::Conv_720[FLOAT, 112x112x1x1]
%onnx::Conv_723[FLOAT, 112x112x1x1]
%onnx::Conv_726[FLOAT, 112x1x3x3]
%onnx::Conv_729[FLOAT, 184x112x1x1]
%onnx::Conv_730[FLOAT, 184]
%onnx::Conv_732[FLOAT, 184x184x1x1]
%onnx::Conv_735[FLOAT, 184x1x5x5]
%onnx::Conv_738[FLOAT, 184x184x1x1]
%onnx::Conv_741[FLOAT, 1104x184x1x1]
%onnx::Conv_742[FLOAT, 1104]
%onnx::Conv_744[FLOAT, 1104x1x3x3]
%onnx::Conv_747[FLOAT, 352x1104x1x1]
%onnx::Conv_748[FLOAT, 352]
%onnx::Conv_750[FLOAT, 1504x352x1x1]
%onnx::Conv_751[FLOAT, 1504]
) {
%onnx::Conv_745 = Identity(%onnx::Conv_742)
%onnx::Conv_739 = Identity(%onnx::Conv_730)
%onnx::Conv_736 = Identity(%onnx::Conv_730)
%onnx::Conv_733 = Identity(%onnx::Conv_730)
%onnx::Conv_727 = Identity(%onnx::Conv_712)
%onnx::Conv_724 = Identity(%onnx::Conv_712)
%onnx::Conv_721 = Identity(%onnx::Conv_712)
%onnx::Conv_718 = Identity(%onnx::Conv_712)
%onnx::Conv_715 = Identity(%onnx::Conv_712)
%onnx::Conv_709 = Identity(%onnx::Conv_676)
%onnx::Conv_706 = Identity(%onnx::Conv_676)
%onnx::Conv_703 = Identity(%onnx::Conv_676)
%onnx::Conv_700 = Identity(%onnx::Conv_676)
%onnx::Conv_697 = Identity(%onnx::Conv_676)
%onnx::Conv_694 = Identity(%onnx::Conv_676)
%onnx::Conv_691 = Identity(%onnx::Conv_676)
%onnx::Conv_688 = Identity(%onnx::Conv_676)
%onnx::Conv_685 = Identity(%onnx::Conv_676)
%onnx::Conv_682 = Identity(%onnx::Conv_670)
%onnx::Conv_679 = Identity(%onnx::Conv_670)
%onnx::Conv_673 = Identity(%onnx::Conv_670)
%onnx::Conv_667 = Identity(%onnx::Conv_640)
%onnx::Conv_664 = Identity(%onnx::Conv_640)
%onnx::Conv_661 = Identity(%onnx::Conv_640)
%onnx::Conv_658 = Identity(%onnx::Conv_640)
%onnx::Conv_655 = Identity(%onnx::Conv_598)
%onnx::Conv_652 = Identity(%onnx::Conv_598)
%onnx::Conv_649 = Identity(%onnx::Conv_640)
%onnx::Conv_646 = Identity(%onnx::Conv_640)
%onnx::Conv_643 = Identity(%onnx::Conv_640)
%onnx::Conv_637 = Identity(%onnx::Conv_616)
%onnx::Conv_634 = Identity(%onnx::Conv_616)
%onnx::Conv_631 = Identity(%onnx::Conv_604)
%onnx::Conv_628 = Identity(%onnx::Conv_616)
%onnx::Conv_625 = Identity(%onnx::Conv_616)
%onnx::Conv_622 = Identity(%onnx::Conv_604)
%onnx::Conv_619 = Identity(%onnx::Conv_616)
%onnx::Conv_613 = Identity(%onnx::Conv_604)
%onnx::Conv_610 = Identity(%onnx::Conv_607)
%onnx::Conv_601 = Identity(%onnx::Conv_598)
%onnx::Conv_595 = Identity(%onnx::Conv_586)
%onnx::Conv_592 = Identity(%onnx::Conv_589)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_585, %onnx::Conv_586)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_588, %onnx::Conv_589)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_591, %onnx::Conv_592)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_594, %onnx::Conv_595)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_597, %onnx::Conv_598)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_600, %onnx::Conv_601)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_603, %onnx::Conv_604)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_606, %onnx::Conv_607)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_609, %onnx::Conv_610)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_612, %onnx::Conv_613)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_615, %onnx::Conv_616)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_618, %onnx::Conv_619)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_621, %onnx::Conv_622)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_624, %onnx::Conv_625)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_627, %onnx::Conv_628)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_630, %onnx::Conv_631)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%583 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %583
}
|
val_accuracy
| 0
| 77,547,648
| 1,566,276
|
{'zcp_synflow': 70.01250983034684, 'zcp_zen': 59.097686767578125, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.09493882209062576, 'zcp_flops': 77547648.0, 'zcp_grad_norm': 23.690643310546875, 'zcp_grasp': -0.016834259033203125, 'zcp_jacov': -16.04866389609505, 'zcp_l2_norm': 503.9791564941406, 'zcp_nwot': 218.43961030356297, 'zcp_params': 1566276.0, 'zcp_plain': -0.004820580128580332, 'zcp_snip': 39.483734130859375, 'lat_1080ti_1': 0.45250835961239755, 'lat_1080ti_32': 0.6738170195004372, 'lat_1080ti_64': 0.7946511967377726, 'lat_2080ti_1': 0.4182169579179831, 'lat_2080ti_32': 0.6387067760809579, 'lat_2080ti_64': 0.7800039337553871, 'lat_essential_ph_1': 0.2830188679245283, 'lat_eyeriss': 0.6157831038087028, 'lat_fpga': 0.4619083901483292, 'lat_gold_6226': 0.24532450120644522, 'lat_gold_6240': 0.22612902863957188, 'lat_pixel2': 0.2391304347826087, 'lat_pixel3': 0.7143666791209397, 'lat_raspi4': 0.7399597214918018, 'lat_samsung_a50': 0.17894736842105263, 'lat_samsung_s7': 0.4015748031496063, 'lat_silver_4114': 0.25242357773065277, 'lat_silver_4210r': 0.2362130120452911, 'lat_titan_rtx_1': 0.3785917670582257, 'lat_titan_rtx_32': 0.5445622717354887, 'lat_titan_rtx_64': 0.7505565435222578, 'lat_titanx_1': 0.20015217676243002, 'lat_titanx_32': 0.7066806633911782, 'lat_titanx_64': 0.7560491482855443, 'lat_titanxp_1': 0.3905968229819633, 'lat_titanxp_32': 0.6578318605745636, 'lat_titanxp_64': 0.7975166002489262}
| |
FBNet_4405
|
FBNet
|
4405
|
4405
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_542[FLOAT, 16x3x3x3]
%onnx::Conv_543[FLOAT, 16]
%onnx::Conv_545[FLOAT, 48x16x1x1]
%onnx::Conv_546[FLOAT, 48]
%onnx::Conv_548[FLOAT, 48x1x3x3]
%onnx::Conv_551[FLOAT, 16x48x1x1]
%onnx::Conv_554[FLOAT, 96x16x1x1]
%onnx::Conv_555[FLOAT, 96]
%onnx::Conv_557[FLOAT, 96x1x3x3]
%onnx::Conv_560[FLOAT, 24x96x1x1]
%onnx::Conv_561[FLOAT, 24]
%onnx::Conv_563[FLOAT, 144x24x1x1]
%onnx::Conv_564[FLOAT, 144]
%onnx::Conv_566[FLOAT, 144x1x5x5]
%onnx::Conv_569[FLOAT, 24x144x1x1]
%onnx::Conv_572[FLOAT, 72x24x1x1]
%onnx::Conv_573[FLOAT, 72]
%onnx::Conv_575[FLOAT, 72x1x3x3]
%onnx::Conv_578[FLOAT, 24x72x1x1]
%onnx::Conv_581[FLOAT, 24x12x1x1]
%onnx::Conv_584[FLOAT, 24x1x3x3]
%onnx::Conv_587[FLOAT, 24x12x1x1]
%onnx::Conv_590[FLOAT, 72x24x1x1]
%onnx::Conv_593[FLOAT, 72x1x5x5]
%onnx::Conv_596[FLOAT, 32x72x1x1]
%onnx::Conv_597[FLOAT, 32]
%onnx::Conv_599[FLOAT, 96x32x1x1]
%onnx::Conv_602[FLOAT, 96x1x3x3]
%onnx::Conv_605[FLOAT, 32x96x1x1]
%onnx::Conv_608[FLOAT, 192x32x1x1]
%onnx::Conv_609[FLOAT, 192]
%onnx::Conv_611[FLOAT, 192x1x3x3]
%onnx::Conv_614[FLOAT, 32x192x1x1]
%onnx::Conv_617[FLOAT, 32x32x1x1]
%onnx::Conv_620[FLOAT, 32x1x3x3]
%onnx::Conv_623[FLOAT, 32x32x1x1]
%onnx::Conv_626[FLOAT, 64x32x1x1]
%onnx::Conv_627[FLOAT, 64]
%onnx::Conv_629[FLOAT, 384x64x1x1]
%onnx::Conv_630[FLOAT, 384]
%onnx::Conv_632[FLOAT, 384x1x5x5]
%onnx::Conv_635[FLOAT, 64x384x1x1]
%onnx::Conv_638[FLOAT, 192x64x1x1]
%onnx::Conv_641[FLOAT, 192x1x3x3]
%onnx::Conv_644[FLOAT, 64x192x1x1]
%onnx::Conv_647[FLOAT, 192x64x1x1]
%onnx::Conv_650[FLOAT, 192x1x3x3]
%onnx::Conv_653[FLOAT, 64x192x1x1]
%onnx::Conv_656[FLOAT, 112x64x1x1]
%onnx::Conv_657[FLOAT, 112]
%onnx::Conv_659[FLOAT, 112x112x1x1]
%onnx::Conv_662[FLOAT, 112x1x5x5]
%onnx::Conv_665[FLOAT, 112x112x1x1]
%onnx::Conv_668[FLOAT, 112x56x1x1]
%onnx::Conv_671[FLOAT, 112x1x3x3]
%onnx::Conv_674[FLOAT, 112x56x1x1]
%onnx::Conv_677[FLOAT, 336x112x1x1]
%onnx::Conv_678[FLOAT, 336]
%onnx::Conv_680[FLOAT, 336x1x5x5]
%onnx::Conv_683[FLOAT, 184x336x1x1]
%onnx::Conv_684[FLOAT, 184]
%onnx::Conv_686[FLOAT, 1104x184x1x1]
%onnx::Conv_687[FLOAT, 1104]
%onnx::Conv_689[FLOAT, 1104x1x5x5]
%onnx::Conv_692[FLOAT, 184x1104x1x1]
%onnx::Conv_695[FLOAT, 1104x184x1x1]
%onnx::Conv_698[FLOAT, 1104x1x5x5]
%onnx::Conv_701[FLOAT, 352x1104x1x1]
%onnx::Conv_702[FLOAT, 352]
%onnx::Conv_704[FLOAT, 1504x352x1x1]
%onnx::Conv_705[FLOAT, 1504]
) {
%onnx::Conv_699 = Identity(%onnx::Conv_687)
%onnx::Conv_696 = Identity(%onnx::Conv_687)
%onnx::Conv_693 = Identity(%onnx::Conv_684)
%onnx::Conv_690 = Identity(%onnx::Conv_687)
%onnx::Conv_681 = Identity(%onnx::Conv_678)
%onnx::Conv_675 = Identity(%onnx::Conv_657)
%onnx::Conv_672 = Identity(%onnx::Conv_657)
%onnx::Conv_669 = Identity(%onnx::Conv_657)
%onnx::Conv_666 = Identity(%onnx::Conv_657)
%onnx::Conv_663 = Identity(%onnx::Conv_657)
%onnx::Conv_660 = Identity(%onnx::Conv_657)
%onnx::Conv_654 = Identity(%onnx::Conv_627)
%onnx::Conv_651 = Identity(%onnx::Conv_609)
%onnx::Conv_648 = Identity(%onnx::Conv_609)
%onnx::Conv_645 = Identity(%onnx::Conv_627)
%onnx::Conv_642 = Identity(%onnx::Conv_609)
%onnx::Conv_639 = Identity(%onnx::Conv_609)
%onnx::Conv_636 = Identity(%onnx::Conv_627)
%onnx::Conv_633 = Identity(%onnx::Conv_630)
%onnx::Conv_624 = Identity(%onnx::Conv_597)
%onnx::Conv_621 = Identity(%onnx::Conv_597)
%onnx::Conv_618 = Identity(%onnx::Conv_597)
%onnx::Conv_615 = Identity(%onnx::Conv_597)
%onnx::Conv_612 = Identity(%onnx::Conv_609)
%onnx::Conv_606 = Identity(%onnx::Conv_597)
%onnx::Conv_603 = Identity(%onnx::Conv_555)
%onnx::Conv_600 = Identity(%onnx::Conv_555)
%onnx::Conv_594 = Identity(%onnx::Conv_573)
%onnx::Conv_591 = Identity(%onnx::Conv_573)
%onnx::Conv_588 = Identity(%onnx::Conv_561)
%onnx::Conv_585 = Identity(%onnx::Conv_561)
%onnx::Conv_582 = Identity(%onnx::Conv_561)
%onnx::Conv_579 = Identity(%onnx::Conv_561)
%onnx::Conv_576 = Identity(%onnx::Conv_573)
%onnx::Conv_570 = Identity(%onnx::Conv_561)
%onnx::Conv_567 = Identity(%onnx::Conv_564)
%onnx::Conv_558 = Identity(%onnx::Conv_555)
%onnx::Conv_552 = Identity(%onnx::Conv_543)
%onnx::Conv_549 = Identity(%onnx::Conv_546)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_542, %onnx::Conv_543)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_545, %onnx::Conv_546)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_548, %onnx::Conv_549)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_551, %onnx::Conv_552)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_554, %onnx::Conv_555)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_557, %onnx::Conv_558)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_560, %onnx::Conv_561)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_563, %onnx::Conv_564)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_566, %onnx::Conv_567)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_569, %onnx::Conv_570)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_572, %onnx::Conv_573)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_575, %onnx::Conv_576)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_578, %onnx::Conv_579)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_581, %onnx::Conv_582)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_584, %onnx::Conv_585)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_587, %onnx::Conv_588)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_590, %onnx::Conv_591)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_593, %onnx::Conv_594)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_596, %onnx::Conv_597)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_599, %onnx::Conv_600)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_602, %onnx::Conv_603)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_605, %onnx::Conv_606)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_608, %onnx::Conv_609)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_611, %onnx::Conv_612)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_614, %onnx::Conv_615)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_617, %onnx::Conv_618)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_620, %onnx::Conv_621)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_623, %onnx::Conv_624)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.8/Add_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.9/relu/Relu_output_0 = Relu(%/cells.9/conv/Conv_output_0)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/relu/Relu_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/relu/Relu_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.13/relu/Relu_output_0 = Relu(%/cells.13/conv/Conv_output_0)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/relu/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/relu/Relu_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%540 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %540
}
|
val_accuracy
| 0
| 74,908,032
| 2,078,092
|
{'zcp_synflow': 71.4280185430511, 'zcp_zen': 60.32728576660156, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.08943033218383789, 'zcp_flops': 74908032.0, 'zcp_grad_norm': 21.239578247070312, 'zcp_grasp': 0.02089691162109375, 'zcp_jacov': -16.05810432054765, 'zcp_l2_norm': 568.9418334960938, 'zcp_nwot': 216.11723543473218, 'zcp_params': 2078092.0, 'zcp_plain': -0.002411236986517906, 'zcp_snip': 38.645545959472656, 'lat_1080ti_1': 0.41201046761493026, 'lat_1080ti_32': 0.37784775830107703, 'lat_1080ti_64': 0.4787169087134528, 'lat_2080ti_1': 0.35884530363445744, 'lat_2080ti_32': 0.4107901758003709, 'lat_2080ti_64': 0.47169940649457875, 'lat_essential_ph_1': 0.18867924528301888, 'lat_eyeriss': 0.5627234885298806, 'lat_fpga': 0.5318279420932327, 'lat_gold_6226': 0.40619809226937065, 'lat_gold_6240': 0.38028505683267194, 'lat_pixel2': 0.4782608695652174, 'lat_pixel3': 0.5440892261183473, 'lat_raspi4': 0.5941367967515916, 'lat_samsung_a50': 0.22105263157894736, 'lat_samsung_s7': 0.15748031496062992, 'lat_silver_4114': 0.42591748244073563, 'lat_silver_4210r': 0.3876508585135945, 'lat_titan_rtx_1': 0.32092667576862094, 'lat_titan_rtx_32': 0.3670992905740317, 'lat_titan_rtx_64': 0.4262059433701156, 'lat_titanx_1': 0.17995301009320064, 'lat_titanx_32': 0.3914593174632346, 'lat_titanx_64': 0.49123823866160315, 'lat_titanxp_1': 0.29576780372999384, 'lat_titanxp_32': 0.3964224330698364, 'lat_titanxp_64': 0.4665916429854239}
| |
FBNet_4063
|
FBNet
|
4063
|
4063
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_723[FLOAT, 16x3x3x3]
%onnx::Conv_724[FLOAT, 16]
%onnx::Conv_726[FLOAT, 48x16x1x1]
%onnx::Conv_727[FLOAT, 48]
%onnx::Conv_729[FLOAT, 48x1x3x3]
%onnx::Conv_732[FLOAT, 16x48x1x1]
%onnx::Conv_735[FLOAT, 16x8x1x1]
%onnx::Conv_738[FLOAT, 16x1x3x3]
%onnx::Conv_741[FLOAT, 24x8x1x1]
%onnx::Conv_742[FLOAT, 24]
%onnx::Conv_744[FLOAT, 72x24x1x1]
%onnx::Conv_745[FLOAT, 72]
%onnx::Conv_747[FLOAT, 72x1x5x5]
%onnx::Conv_750[FLOAT, 24x72x1x1]
%onnx::Conv_753[FLOAT, 24x12x1x1]
%onnx::Conv_756[FLOAT, 24x1x3x3]
%onnx::Conv_759[FLOAT, 24x12x1x1]
%onnx::Conv_762[FLOAT, 144x24x1x1]
%onnx::Conv_763[FLOAT, 144]
%onnx::Conv_765[FLOAT, 144x1x3x3]
%onnx::Conv_768[FLOAT, 32x144x1x1]
%onnx::Conv_769[FLOAT, 32]
%onnx::Conv_771[FLOAT, 192x32x1x1]
%onnx::Conv_772[FLOAT, 192]
%onnx::Conv_774[FLOAT, 192x1x5x5]
%onnx::Conv_777[FLOAT, 32x192x1x1]
%onnx::Conv_780[FLOAT, 32x32x1x1]
%onnx::Conv_783[FLOAT, 32x1x5x5]
%onnx::Conv_786[FLOAT, 32x32x1x1]
%onnx::Conv_789[FLOAT, 32x16x1x1]
%onnx::Conv_792[FLOAT, 32x1x5x5]
%onnx::Conv_795[FLOAT, 32x16x1x1]
%onnx::Conv_798[FLOAT, 32x32x1x1]
%onnx::Conv_801[FLOAT, 32x1x5x5]
%onnx::Conv_804[FLOAT, 64x32x1x1]
%onnx::Conv_805[FLOAT, 64]
%onnx::Conv_807[FLOAT, 64x32x1x1]
%onnx::Conv_810[FLOAT, 64x1x5x5]
%onnx::Conv_813[FLOAT, 64x32x1x1]
%onnx::Conv_816[FLOAT, 64x32x1x1]
%onnx::Conv_819[FLOAT, 64x1x5x5]
%onnx::Conv_822[FLOAT, 64x32x1x1]
%onnx::Conv_825[FLOAT, 64x64x1x1]
%onnx::Conv_828[FLOAT, 64x1x3x3]
%onnx::Conv_831[FLOAT, 64x64x1x1]
%onnx::Conv_834[FLOAT, 64x32x1x1]
%onnx::Conv_837[FLOAT, 64x1x3x3]
%onnx::Conv_840[FLOAT, 112x32x1x1]
%onnx::Conv_841[FLOAT, 112]
%onnx::Conv_843[FLOAT, 112x112x1x1]
%onnx::Conv_846[FLOAT, 112x1x3x3]
%onnx::Conv_849[FLOAT, 112x112x1x1]
%onnx::Conv_852[FLOAT, 672x112x1x1]
%onnx::Conv_853[FLOAT, 672]
%onnx::Conv_855[FLOAT, 672x1x5x5]
%onnx::Conv_858[FLOAT, 112x672x1x1]
%onnx::Conv_861[FLOAT, 672x112x1x1]
%onnx::Conv_864[FLOAT, 672x1x5x5]
%onnx::Conv_867[FLOAT, 112x672x1x1]
%onnx::Conv_870[FLOAT, 672x112x1x1]
%onnx::Conv_873[FLOAT, 672x1x5x5]
%onnx::Conv_876[FLOAT, 184x672x1x1]
%onnx::Conv_877[FLOAT, 184]
%onnx::Conv_879[FLOAT, 552x184x1x1]
%onnx::Conv_880[FLOAT, 552]
%onnx::Conv_882[FLOAT, 552x1x5x5]
%onnx::Conv_885[FLOAT, 184x552x1x1]
%onnx::Conv_888[FLOAT, 184x184x1x1]
%onnx::Conv_891[FLOAT, 184x1x5x5]
%onnx::Conv_894[FLOAT, 184x184x1x1]
%onnx::Conv_897[FLOAT, 1104x184x1x1]
%onnx::Conv_898[FLOAT, 1104]
%onnx::Conv_900[FLOAT, 1104x1x5x5]
%onnx::Conv_903[FLOAT, 184x1104x1x1]
%onnx::Conv_906[FLOAT, 184x92x1x1]
%onnx::Conv_909[FLOAT, 184x1x5x5]
%onnx::Conv_912[FLOAT, 352x92x1x1]
%onnx::Conv_913[FLOAT, 352]
%onnx::Conv_915[FLOAT, 1504x352x1x1]
%onnx::Conv_916[FLOAT, 1504]
) {
%onnx::Conv_910 = Identity(%onnx::Conv_877)
%onnx::Conv_907 = Identity(%onnx::Conv_877)
%onnx::Conv_904 = Identity(%onnx::Conv_877)
%onnx::Conv_901 = Identity(%onnx::Conv_898)
%onnx::Conv_895 = Identity(%onnx::Conv_877)
%onnx::Conv_892 = Identity(%onnx::Conv_877)
%onnx::Conv_889 = Identity(%onnx::Conv_877)
%onnx::Conv_886 = Identity(%onnx::Conv_877)
%onnx::Conv_883 = Identity(%onnx::Conv_880)
%onnx::Conv_874 = Identity(%onnx::Conv_853)
%onnx::Conv_871 = Identity(%onnx::Conv_853)
%onnx::Conv_868 = Identity(%onnx::Conv_841)
%onnx::Conv_865 = Identity(%onnx::Conv_853)
%onnx::Conv_862 = Identity(%onnx::Conv_853)
%onnx::Conv_859 = Identity(%onnx::Conv_841)
%onnx::Conv_856 = Identity(%onnx::Conv_853)
%onnx::Conv_850 = Identity(%onnx::Conv_841)
%onnx::Conv_847 = Identity(%onnx::Conv_841)
%onnx::Conv_844 = Identity(%onnx::Conv_841)
%onnx::Conv_838 = Identity(%onnx::Conv_805)
%onnx::Conv_835 = Identity(%onnx::Conv_805)
%onnx::Conv_832 = Identity(%onnx::Conv_805)
%onnx::Conv_829 = Identity(%onnx::Conv_805)
%onnx::Conv_826 = Identity(%onnx::Conv_805)
%onnx::Conv_823 = Identity(%onnx::Conv_805)
%onnx::Conv_820 = Identity(%onnx::Conv_805)
%onnx::Conv_817 = Identity(%onnx::Conv_805)
%onnx::Conv_814 = Identity(%onnx::Conv_805)
%onnx::Conv_811 = Identity(%onnx::Conv_805)
%onnx::Conv_808 = Identity(%onnx::Conv_805)
%onnx::Conv_802 = Identity(%onnx::Conv_769)
%onnx::Conv_799 = Identity(%onnx::Conv_769)
%onnx::Conv_796 = Identity(%onnx::Conv_769)
%onnx::Conv_793 = Identity(%onnx::Conv_769)
%onnx::Conv_790 = Identity(%onnx::Conv_769)
%onnx::Conv_787 = Identity(%onnx::Conv_769)
%onnx::Conv_784 = Identity(%onnx::Conv_769)
%onnx::Conv_781 = Identity(%onnx::Conv_769)
%onnx::Conv_778 = Identity(%onnx::Conv_769)
%onnx::Conv_775 = Identity(%onnx::Conv_772)
%onnx::Conv_766 = Identity(%onnx::Conv_763)
%onnx::Conv_760 = Identity(%onnx::Conv_742)
%onnx::Conv_757 = Identity(%onnx::Conv_742)
%onnx::Conv_754 = Identity(%onnx::Conv_742)
%onnx::Conv_751 = Identity(%onnx::Conv_742)
%onnx::Conv_748 = Identity(%onnx::Conv_745)
%onnx::Conv_739 = Identity(%onnx::Conv_724)
%onnx::Conv_736 = Identity(%onnx::Conv_724)
%onnx::Conv_733 = Identity(%onnx::Conv_724)
%onnx::Conv_730 = Identity(%onnx::Conv_727)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_723, %onnx::Conv_724)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_858, %onnx::Conv_859)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_861, %onnx::Conv_862)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_864, %onnx::Conv_865)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_867, %onnx::Conv_868)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_870, %onnx::Conv_871)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_873, %onnx::Conv_874)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_876, %onnx::Conv_877)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_879, %onnx::Conv_880)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_882, %onnx::Conv_883)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_885, %onnx::Conv_886)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_888, %onnx::Conv_889)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_891, %onnx::Conv_892)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_894, %onnx::Conv_895)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_897, %onnx::Conv_898)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_900, %onnx::Conv_901)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_903, %onnx::Conv_904)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_906, %onnx::Conv_907)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_909, %onnx::Conv_910)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_912, %onnx::Conv_913)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_915, %onnx::Conv_916)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%721 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %721
}
|
val_accuracy
| 0
| 76,003,456
| 2,130,732
|
{'zcp_synflow': 77.77928464665416, 'zcp_zen': 69.78970336914062, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.1202029287815094, 'zcp_flops': 76003456.0, 'zcp_grad_norm': 22.43703269958496, 'zcp_grasp': 0.2164764404296875, 'zcp_jacov': -16.06525458846213, 'zcp_l2_norm': 645.6082153320312, 'zcp_nwot': 211.9324341074611, 'zcp_params': 2130732.0, 'zcp_plain': -0.003136718412861228, 'zcp_snip': 40.03001022338867, 'lat_1080ti_1': 0.7623304760037224, 'lat_1080ti_32': 0.6446969614126391, 'lat_1080ti_64': 0.49447931498158587, 'lat_2080ti_1': 0.7935130004746421, 'lat_2080ti_32': 0.6427726990470344, 'lat_2080ti_64': 0.5476655079376612, 'lat_essential_ph_1': 0.37735849056603776, 'lat_eyeriss': 0.4981387416793501, 'lat_fpga': 0.5112162202277876, 'lat_gold_6226': 0.5841952451224302, 'lat_gold_6240': 0.557338832925976, 'lat_pixel2': 0.34782608695652173, 'lat_pixel3': 0.5267027292325931, 'lat_raspi4': 0.5005715025116204, 'lat_samsung_a50': 0.24210526315789474, 'lat_samsung_s7': 0.2047244094488189, 'lat_silver_4114': 0.6398914873519249, 'lat_silver_4210r': 0.6728366351980954, 'lat_titan_rtx_1': 0.7390386492347728, 'lat_titan_rtx_32': 0.6137212694434767, 'lat_titan_rtx_64': 0.5681692694536806, 'lat_titanx_1': 0.3961883858748438, 'lat_titanx_32': 0.5706776994435363, 'lat_titanx_64': 0.5384404832009533, 'lat_titanxp_1': 0.697047418136899, 'lat_titanxp_32': 0.5871612133900049, 'lat_titanxp_64': 0.5228677300197793}
| |
FBNet_3311
|
FBNet
|
3311
|
3311
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_569[FLOAT, 16x3x3x3]
%onnx::Conv_570[FLOAT, 16]
%onnx::Conv_572[FLOAT, 16x16x1x1]
%onnx::Conv_575[FLOAT, 16x1x5x5]
%onnx::Conv_578[FLOAT, 24x16x1x1]
%onnx::Conv_579[FLOAT, 24]
%onnx::Conv_581[FLOAT, 24x24x1x1]
%onnx::Conv_584[FLOAT, 24x1x3x3]
%onnx::Conv_587[FLOAT, 24x24x1x1]
%onnx::Conv_590[FLOAT, 24x12x1x1]
%onnx::Conv_593[FLOAT, 24x1x3x3]
%onnx::Conv_596[FLOAT, 24x12x1x1]
%onnx::Conv_599[FLOAT, 24x12x1x1]
%onnx::Conv_602[FLOAT, 24x1x3x3]
%onnx::Conv_605[FLOAT, 24x12x1x1]
%onnx::Conv_608[FLOAT, 72x24x1x1]
%onnx::Conv_609[FLOAT, 72]
%onnx::Conv_611[FLOAT, 72x1x3x3]
%onnx::Conv_614[FLOAT, 32x72x1x1]
%onnx::Conv_615[FLOAT, 32]
%onnx::Conv_617[FLOAT, 32x16x1x1]
%onnx::Conv_620[FLOAT, 32x1x5x5]
%onnx::Conv_623[FLOAT, 32x16x1x1]
%onnx::Conv_626[FLOAT, 192x32x1x1]
%onnx::Conv_627[FLOAT, 192]
%onnx::Conv_629[FLOAT, 192x1x5x5]
%onnx::Conv_632[FLOAT, 32x192x1x1]
%onnx::Conv_635[FLOAT, 32x32x1x1]
%onnx::Conv_638[FLOAT, 32x1x5x5]
%onnx::Conv_641[FLOAT, 64x32x1x1]
%onnx::Conv_642[FLOAT, 64]
%onnx::Conv_644[FLOAT, 384x64x1x1]
%onnx::Conv_645[FLOAT, 384]
%onnx::Conv_647[FLOAT, 384x1x3x3]
%onnx::Conv_650[FLOAT, 64x384x1x1]
%onnx::Conv_653[FLOAT, 192x64x1x1]
%onnx::Conv_656[FLOAT, 192x1x5x5]
%onnx::Conv_659[FLOAT, 64x192x1x1]
%onnx::Conv_662[FLOAT, 64x64x1x1]
%onnx::Conv_665[FLOAT, 64x1x3x3]
%onnx::Conv_668[FLOAT, 64x64x1x1]
%onnx::Conv_671[FLOAT, 64x64x1x1]
%onnx::Conv_674[FLOAT, 64x1x5x5]
%onnx::Conv_677[FLOAT, 112x64x1x1]
%onnx::Conv_678[FLOAT, 112]
%onnx::Conv_680[FLOAT, 112x56x1x1]
%onnx::Conv_683[FLOAT, 112x1x3x3]
%onnx::Conv_686[FLOAT, 112x56x1x1]
%onnx::Conv_689[FLOAT, 112x112x1x1]
%onnx::Conv_692[FLOAT, 112x1x3x3]
%onnx::Conv_695[FLOAT, 112x112x1x1]
%onnx::Conv_698[FLOAT, 112x112x1x1]
%onnx::Conv_701[FLOAT, 112x1x3x3]
%onnx::Conv_704[FLOAT, 112x112x1x1]
%onnx::Conv_707[FLOAT, 112x112x1x1]
%onnx::Conv_710[FLOAT, 112x1x3x3]
%onnx::Conv_713[FLOAT, 184x112x1x1]
%onnx::Conv_714[FLOAT, 184]
%onnx::Conv_716[FLOAT, 552x184x1x1]
%onnx::Conv_717[FLOAT, 552]
%onnx::Conv_719[FLOAT, 552x1x3x3]
%onnx::Conv_722[FLOAT, 184x552x1x1]
%onnx::Conv_725[FLOAT, 352x184x1x1]
%onnx::Conv_726[FLOAT, 352]
%onnx::Conv_728[FLOAT, 1504x352x1x1]
%onnx::Conv_729[FLOAT, 1504]
) {
%onnx::Conv_723 = Identity(%onnx::Conv_714)
%onnx::Conv_720 = Identity(%onnx::Conv_717)
%onnx::Conv_711 = Identity(%onnx::Conv_678)
%onnx::Conv_708 = Identity(%onnx::Conv_678)
%onnx::Conv_705 = Identity(%onnx::Conv_678)
%onnx::Conv_702 = Identity(%onnx::Conv_678)
%onnx::Conv_699 = Identity(%onnx::Conv_678)
%onnx::Conv_696 = Identity(%onnx::Conv_678)
%onnx::Conv_693 = Identity(%onnx::Conv_678)
%onnx::Conv_690 = Identity(%onnx::Conv_678)
%onnx::Conv_687 = Identity(%onnx::Conv_678)
%onnx::Conv_684 = Identity(%onnx::Conv_678)
%onnx::Conv_681 = Identity(%onnx::Conv_678)
%onnx::Conv_675 = Identity(%onnx::Conv_642)
%onnx::Conv_672 = Identity(%onnx::Conv_642)
%onnx::Conv_669 = Identity(%onnx::Conv_642)
%onnx::Conv_666 = Identity(%onnx::Conv_642)
%onnx::Conv_663 = Identity(%onnx::Conv_642)
%onnx::Conv_660 = Identity(%onnx::Conv_642)
%onnx::Conv_657 = Identity(%onnx::Conv_627)
%onnx::Conv_654 = Identity(%onnx::Conv_627)
%onnx::Conv_651 = Identity(%onnx::Conv_642)
%onnx::Conv_648 = Identity(%onnx::Conv_645)
%onnx::Conv_639 = Identity(%onnx::Conv_615)
%onnx::Conv_636 = Identity(%onnx::Conv_615)
%onnx::Conv_633 = Identity(%onnx::Conv_615)
%onnx::Conv_630 = Identity(%onnx::Conv_627)
%onnx::Conv_624 = Identity(%onnx::Conv_615)
%onnx::Conv_621 = Identity(%onnx::Conv_615)
%onnx::Conv_618 = Identity(%onnx::Conv_615)
%onnx::Conv_612 = Identity(%onnx::Conv_609)
%onnx::Conv_606 = Identity(%onnx::Conv_579)
%onnx::Conv_603 = Identity(%onnx::Conv_579)
%onnx::Conv_600 = Identity(%onnx::Conv_579)
%onnx::Conv_597 = Identity(%onnx::Conv_579)
%onnx::Conv_594 = Identity(%onnx::Conv_579)
%onnx::Conv_591 = Identity(%onnx::Conv_579)
%onnx::Conv_588 = Identity(%onnx::Conv_579)
%onnx::Conv_585 = Identity(%onnx::Conv_579)
%onnx::Conv_582 = Identity(%onnx::Conv_579)
%onnx::Conv_576 = Identity(%onnx::Conv_570)
%onnx::Conv_573 = Identity(%onnx::Conv_570)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_569, %onnx::Conv_570)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_572, %onnx::Conv_573)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_575, %onnx::Conv_576)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_578, %onnx::Conv_579)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_581, %onnx::Conv_582)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_584, %onnx::Conv_585)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_587, %onnx::Conv_588)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_590, %onnx::Conv_591)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_593, %onnx::Conv_594)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_596, %onnx::Conv_597)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_599, %onnx::Conv_600)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_602, %onnx::Conv_603)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_605, %onnx::Conv_606)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_608, %onnx::Conv_609)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_611, %onnx::Conv_612)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_614, %onnx::Conv_615)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_617, %onnx::Conv_618)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_620, %onnx::Conv_621)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_623, %onnx::Conv_624)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.21/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.21/relu/Relu_output_0 = Relu(%/cells.21/conv/Conv_output_0)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/relu/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%567 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %567
}
|
val_accuracy
| 0
| 39,185,024
| 1,202,940
|
{'zcp_synflow': 67.89484470061731, 'zcp_zen': 55.4381103515625, 'zcp_epe_nas': 8.201088348286222, 'zcp_fisher': 0.03457094728946686, 'zcp_flops': 39185024.0, 'zcp_grad_norm': 13.870241165161133, 'zcp_grasp': 0.004350423812866211, 'zcp_jacov': -16.052256879447047, 'zcp_l2_norm': 470.15966796875, 'zcp_nwot': 203.37304559337477, 'zcp_params': 1202940.0, 'zcp_plain': -0.0013056622119620442, 'zcp_snip': 21.38772201538086, 'lat_1080ti_1': 0.33135673858124637, 'lat_1080ti_32': 0.23564093747420237, 'lat_1080ti_64': 0.11348356052367377, 'lat_2080ti_1': 0.3323160031587409, 'lat_2080ti_32': 0.24339636610090568, 'lat_2080ti_64': 0.1395302185588714, 'lat_essential_ph_1': 0.09433962264150944, 'lat_eyeriss': 0.09458493022573491, 'lat_fpga': 0.0734422466084168, 'lat_gold_6226': 0.09093462242545448, 'lat_gold_6240': 0.14070295201102626, 'lat_pixel2': 0.15217391304347827, 'lat_pixel3': 0.10465329678253589, 'lat_raspi4': 0.11428858364368684, 'lat_samsung_a50': 0.031578947368421054, 'lat_samsung_s7': 0.031496062992125984, 'lat_silver_4114': 0.16246099089669286, 'lat_silver_4210r': 0.13880187089936558, 'lat_titan_rtx_1': 0.30502616673839195, 'lat_titan_rtx_32': 0.23876603567987814, 'lat_titan_rtx_64': 0.1335716162720618, 'lat_titanx_1': 0.1557657824107464, 'lat_titanx_32': 0.14349476028012795, 'lat_titanx_64': 0.11496388707321915, 'lat_titanxp_1': 0.28866556154549466, 'lat_titanxp_32': 0.18790035331529076, 'lat_titanxp_64': 0.12552333919498182}
| |
FBNet_568
|
FBNet
|
568
|
568
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_634[FLOAT, 16x3x3x3]
%onnx::Conv_635[FLOAT, 16]
%onnx::Conv_637[FLOAT, 16x16x1x1]
%onnx::Conv_640[FLOAT, 16x1x3x3]
%onnx::Conv_643[FLOAT, 16x16x1x1]
%onnx::Conv_646[FLOAT, 24x16x1x1]
%onnx::Conv_647[FLOAT, 24]
%onnx::Conv_649[FLOAT, 72x24x1x1]
%onnx::Conv_650[FLOAT, 72]
%onnx::Conv_652[FLOAT, 72x1x5x5]
%onnx::Conv_655[FLOAT, 24x72x1x1]
%onnx::Conv_658[FLOAT, 24x12x1x1]
%onnx::Conv_661[FLOAT, 24x1x5x5]
%onnx::Conv_664[FLOAT, 24x12x1x1]
%onnx::Conv_667[FLOAT, 144x24x1x1]
%onnx::Conv_668[FLOAT, 144]
%onnx::Conv_670[FLOAT, 144x1x3x3]
%onnx::Conv_673[FLOAT, 24x144x1x1]
%onnx::Conv_676[FLOAT, 144x24x1x1]
%onnx::Conv_679[FLOAT, 144x1x3x3]
%onnx::Conv_682[FLOAT, 32x144x1x1]
%onnx::Conv_683[FLOAT, 32]
%onnx::Conv_685[FLOAT, 32x32x1x1]
%onnx::Conv_688[FLOAT, 32x1x5x5]
%onnx::Conv_691[FLOAT, 32x32x1x1]
%onnx::Conv_694[FLOAT, 32x16x1x1]
%onnx::Conv_697[FLOAT, 32x1x3x3]
%onnx::Conv_700[FLOAT, 32x16x1x1]
%onnx::Conv_703[FLOAT, 32x32x1x1]
%onnx::Conv_706[FLOAT, 32x1x5x5]
%onnx::Conv_709[FLOAT, 64x32x1x1]
%onnx::Conv_710[FLOAT, 64]
%onnx::Conv_712[FLOAT, 64x64x1x1]
%onnx::Conv_715[FLOAT, 64x1x3x3]
%onnx::Conv_718[FLOAT, 64x64x1x1]
%onnx::Conv_721[FLOAT, 192x64x1x1]
%onnx::Conv_722[FLOAT, 192]
%onnx::Conv_724[FLOAT, 192x1x3x3]
%onnx::Conv_727[FLOAT, 64x192x1x1]
%onnx::Conv_730[FLOAT, 192x64x1x1]
%onnx::Conv_733[FLOAT, 192x1x5x5]
%onnx::Conv_736[FLOAT, 64x192x1x1]
%onnx::Conv_739[FLOAT, 112x64x1x1]
%onnx::Conv_740[FLOAT, 112]
%onnx::Conv_742[FLOAT, 336x112x1x1]
%onnx::Conv_743[FLOAT, 336]
%onnx::Conv_745[FLOAT, 336x1x5x5]
%onnx::Conv_748[FLOAT, 112x336x1x1]
%onnx::Conv_751[FLOAT, 112x56x1x1]
%onnx::Conv_754[FLOAT, 112x1x5x5]
%onnx::Conv_757[FLOAT, 112x56x1x1]
%onnx::Conv_760[FLOAT, 672x112x1x1]
%onnx::Conv_761[FLOAT, 672]
%onnx::Conv_763[FLOAT, 672x1x3x3]
%onnx::Conv_766[FLOAT, 112x672x1x1]
%onnx::Conv_769[FLOAT, 672x112x1x1]
%onnx::Conv_772[FLOAT, 672x1x3x3]
%onnx::Conv_775[FLOAT, 184x672x1x1]
%onnx::Conv_776[FLOAT, 184]
%onnx::Conv_778[FLOAT, 1104x184x1x1]
%onnx::Conv_779[FLOAT, 1104]
%onnx::Conv_781[FLOAT, 1104x1x3x3]
%onnx::Conv_784[FLOAT, 184x1104x1x1]
%onnx::Conv_787[FLOAT, 1104x184x1x1]
%onnx::Conv_790[FLOAT, 1104x1x3x3]
%onnx::Conv_793[FLOAT, 184x1104x1x1]
%onnx::Conv_796[FLOAT, 184x184x1x1]
%onnx::Conv_799[FLOAT, 184x1x3x3]
%onnx::Conv_802[FLOAT, 184x184x1x1]
%onnx::Conv_805[FLOAT, 184x92x1x1]
%onnx::Conv_808[FLOAT, 184x1x5x5]
%onnx::Conv_811[FLOAT, 352x92x1x1]
%onnx::Conv_812[FLOAT, 352]
%onnx::Conv_814[FLOAT, 1504x352x1x1]
%onnx::Conv_815[FLOAT, 1504]
) {
%onnx::Conv_809 = Identity(%onnx::Conv_776)
%onnx::Conv_806 = Identity(%onnx::Conv_776)
%onnx::Conv_803 = Identity(%onnx::Conv_776)
%onnx::Conv_800 = Identity(%onnx::Conv_776)
%onnx::Conv_797 = Identity(%onnx::Conv_776)
%onnx::Conv_794 = Identity(%onnx::Conv_776)
%onnx::Conv_791 = Identity(%onnx::Conv_779)
%onnx::Conv_788 = Identity(%onnx::Conv_779)
%onnx::Conv_785 = Identity(%onnx::Conv_776)
%onnx::Conv_782 = Identity(%onnx::Conv_779)
%onnx::Conv_773 = Identity(%onnx::Conv_761)
%onnx::Conv_770 = Identity(%onnx::Conv_761)
%onnx::Conv_767 = Identity(%onnx::Conv_740)
%onnx::Conv_764 = Identity(%onnx::Conv_761)
%onnx::Conv_758 = Identity(%onnx::Conv_740)
%onnx::Conv_755 = Identity(%onnx::Conv_740)
%onnx::Conv_752 = Identity(%onnx::Conv_740)
%onnx::Conv_749 = Identity(%onnx::Conv_740)
%onnx::Conv_746 = Identity(%onnx::Conv_743)
%onnx::Conv_737 = Identity(%onnx::Conv_710)
%onnx::Conv_734 = Identity(%onnx::Conv_722)
%onnx::Conv_731 = Identity(%onnx::Conv_722)
%onnx::Conv_728 = Identity(%onnx::Conv_710)
%onnx::Conv_725 = Identity(%onnx::Conv_722)
%onnx::Conv_719 = Identity(%onnx::Conv_710)
%onnx::Conv_716 = Identity(%onnx::Conv_710)
%onnx::Conv_713 = Identity(%onnx::Conv_710)
%onnx::Conv_707 = Identity(%onnx::Conv_683)
%onnx::Conv_704 = Identity(%onnx::Conv_683)
%onnx::Conv_701 = Identity(%onnx::Conv_683)
%onnx::Conv_698 = Identity(%onnx::Conv_683)
%onnx::Conv_695 = Identity(%onnx::Conv_683)
%onnx::Conv_692 = Identity(%onnx::Conv_683)
%onnx::Conv_689 = Identity(%onnx::Conv_683)
%onnx::Conv_686 = Identity(%onnx::Conv_683)
%onnx::Conv_680 = Identity(%onnx::Conv_668)
%onnx::Conv_677 = Identity(%onnx::Conv_668)
%onnx::Conv_674 = Identity(%onnx::Conv_647)
%onnx::Conv_671 = Identity(%onnx::Conv_668)
%onnx::Conv_665 = Identity(%onnx::Conv_647)
%onnx::Conv_662 = Identity(%onnx::Conv_647)
%onnx::Conv_659 = Identity(%onnx::Conv_647)
%onnx::Conv_656 = Identity(%onnx::Conv_647)
%onnx::Conv_653 = Identity(%onnx::Conv_650)
%onnx::Conv_644 = Identity(%onnx::Conv_635)
%onnx::Conv_641 = Identity(%onnx::Conv_635)
%onnx::Conv_638 = Identity(%onnx::Conv_635)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_634, %onnx::Conv_635)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_637, %onnx::Conv_638)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_640, %onnx::Conv_641)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_643, %onnx::Conv_644)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_646, %onnx::Conv_647)
%/cells.1/relu/Relu_output_0 = Relu(%/cells.1/conv/Conv_output_0)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/relu/Relu_output_0, %onnx::Conv_649, %onnx::Conv_650)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_652, %onnx::Conv_653)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_655, %onnx::Conv_656)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/relu/Relu_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_658, %onnx::Conv_659)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.13/relu/Relu_output_0 = Relu(%/cells.13/conv/Conv_output_0)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/relu/Relu_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/relu/Relu_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%632 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %632
}
|
val_accuracy
| 0
| 78,030,720
| 2,229,556
|
{'zcp_synflow': 77.13460575703682, 'zcp_zen': 67.40946197509766, 'zcp_epe_nas': 7.748394817578343, 'zcp_fisher': 0.1216965764760971, 'zcp_flops': 78030720.0, 'zcp_grad_norm': 24.13748550415039, 'zcp_grasp': 0.052707672119140625, 'zcp_jacov': -16.06639513346782, 'zcp_l2_norm': 654.1785278320312, 'zcp_nwot': 214.0939572226, 'zcp_params': 2229556.0, 'zcp_plain': 0.004188590217381716, 'zcp_snip': 40.76276397705078, 'lat_1080ti_1': 0.6063468399255849, 'lat_1080ti_32': 0.5101554367529743, 'lat_1080ti_64': 0.46385616876057084, 'lat_2080ti_1': 0.5651329016078172, 'lat_2080ti_32': 0.5618461646771022, 'lat_2080ti_64': 0.4932185551975904, 'lat_essential_ph_1': 0.5094339622641509, 'lat_eyeriss': 0.5069774265123872, 'lat_fpga': 0.5580403526652746, 'lat_gold_6226': 0.4381900046276719, 'lat_gold_6240': 0.6225252606974274, 'lat_pixel2': 0.391304347826087, 'lat_pixel3': 0.48626623410691366, 'lat_raspi4': 0.5432899528675077, 'lat_samsung_a50': 0.23157894736842105, 'lat_samsung_s7': 0.2755905511811024, 'lat_silver_4114': 0.5887558278896887, 'lat_silver_4210r': 0.5959131190831372, 'lat_titan_rtx_1': 0.5482722569411839, 'lat_titan_rtx_32': 0.5212537327498609, 'lat_titan_rtx_64': 0.5056415995981083, 'lat_titanx_1': 0.2933479804065005, 'lat_titanx_32': 0.5327488749263506, 'lat_titanx_64': 0.45069349443734114, 'lat_titanxp_1': 0.5364293803457263, 'lat_titanxp_32': 0.523014125278337, 'lat_titanxp_64': 0.4853400538403634}
| |
FBNet_422
|
FBNet
|
422
|
422
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_578[FLOAT, 16x3x3x3]
%onnx::Conv_579[FLOAT, 16]
%onnx::Conv_581[FLOAT, 96x16x1x1]
%onnx::Conv_582[FLOAT, 96]
%onnx::Conv_584[FLOAT, 96x1x5x5]
%onnx::Conv_587[FLOAT, 16x96x1x1]
%onnx::Conv_590[FLOAT, 48x16x1x1]
%onnx::Conv_591[FLOAT, 48]
%onnx::Conv_593[FLOAT, 48x1x3x3]
%onnx::Conv_596[FLOAT, 24x48x1x1]
%onnx::Conv_597[FLOAT, 24]
%onnx::Conv_599[FLOAT, 144x24x1x1]
%onnx::Conv_600[FLOAT, 144]
%onnx::Conv_602[FLOAT, 144x1x5x5]
%onnx::Conv_605[FLOAT, 24x144x1x1]
%onnx::Conv_608[FLOAT, 24x12x1x1]
%onnx::Conv_611[FLOAT, 24x1x3x3]
%onnx::Conv_614[FLOAT, 24x12x1x1]
%onnx::Conv_617[FLOAT, 24x24x1x1]
%onnx::Conv_620[FLOAT, 24x1x3x3]
%onnx::Conv_623[FLOAT, 32x24x1x1]
%onnx::Conv_624[FLOAT, 32]
%onnx::Conv_626[FLOAT, 96x32x1x1]
%onnx::Conv_629[FLOAT, 96x1x5x5]
%onnx::Conv_632[FLOAT, 32x96x1x1]
%onnx::Conv_635[FLOAT, 96x32x1x1]
%onnx::Conv_638[FLOAT, 96x1x3x3]
%onnx::Conv_641[FLOAT, 32x96x1x1]
%onnx::Conv_644[FLOAT, 32x32x1x1]
%onnx::Conv_647[FLOAT, 32x1x5x5]
%onnx::Conv_650[FLOAT, 32x32x1x1]
%onnx::Conv_653[FLOAT, 192x32x1x1]
%onnx::Conv_654[FLOAT, 192]
%onnx::Conv_656[FLOAT, 192x1x5x5]
%onnx::Conv_659[FLOAT, 64x192x1x1]
%onnx::Conv_660[FLOAT, 64]
%onnx::Conv_662[FLOAT, 192x64x1x1]
%onnx::Conv_665[FLOAT, 192x1x3x3]
%onnx::Conv_668[FLOAT, 64x192x1x1]
%onnx::Conv_671[FLOAT, 192x64x1x1]
%onnx::Conv_674[FLOAT, 192x1x5x5]
%onnx::Conv_677[FLOAT, 64x192x1x1]
%onnx::Conv_680[FLOAT, 192x64x1x1]
%onnx::Conv_683[FLOAT, 192x1x3x3]
%onnx::Conv_686[FLOAT, 112x192x1x1]
%onnx::Conv_687[FLOAT, 112]
%onnx::Conv_689[FLOAT, 112x56x1x1]
%onnx::Conv_692[FLOAT, 112x1x3x3]
%onnx::Conv_695[FLOAT, 112x56x1x1]
%onnx::Conv_698[FLOAT, 672x112x1x1]
%onnx::Conv_699[FLOAT, 672]
%onnx::Conv_701[FLOAT, 672x1x5x5]
%onnx::Conv_704[FLOAT, 112x672x1x1]
%onnx::Conv_707[FLOAT, 672x112x1x1]
%onnx::Conv_710[FLOAT, 672x1x3x3]
%onnx::Conv_713[FLOAT, 112x672x1x1]
%onnx::Conv_716[FLOAT, 184x112x1x1]
%onnx::Conv_717[FLOAT, 184]
%onnx::Conv_719[FLOAT, 184x184x1x1]
%onnx::Conv_722[FLOAT, 184x1x5x5]
%onnx::Conv_725[FLOAT, 184x184x1x1]
%onnx::Conv_728[FLOAT, 184x92x1x1]
%onnx::Conv_731[FLOAT, 184x1x3x3]
%onnx::Conv_734[FLOAT, 184x92x1x1]
%onnx::Conv_737[FLOAT, 552x184x1x1]
%onnx::Conv_738[FLOAT, 552]
%onnx::Conv_740[FLOAT, 552x1x5x5]
%onnx::Conv_743[FLOAT, 352x552x1x1]
%onnx::Conv_744[FLOAT, 352]
%onnx::Conv_746[FLOAT, 1504x352x1x1]
%onnx::Conv_747[FLOAT, 1504]
) {
%onnx::Conv_741 = Identity(%onnx::Conv_738)
%onnx::Conv_735 = Identity(%onnx::Conv_717)
%onnx::Conv_732 = Identity(%onnx::Conv_717)
%onnx::Conv_729 = Identity(%onnx::Conv_717)
%onnx::Conv_726 = Identity(%onnx::Conv_717)
%onnx::Conv_723 = Identity(%onnx::Conv_717)
%onnx::Conv_720 = Identity(%onnx::Conv_717)
%onnx::Conv_714 = Identity(%onnx::Conv_687)
%onnx::Conv_711 = Identity(%onnx::Conv_699)
%onnx::Conv_708 = Identity(%onnx::Conv_699)
%onnx::Conv_705 = Identity(%onnx::Conv_687)
%onnx::Conv_702 = Identity(%onnx::Conv_699)
%onnx::Conv_696 = Identity(%onnx::Conv_687)
%onnx::Conv_693 = Identity(%onnx::Conv_687)
%onnx::Conv_690 = Identity(%onnx::Conv_687)
%onnx::Conv_684 = Identity(%onnx::Conv_654)
%onnx::Conv_681 = Identity(%onnx::Conv_654)
%onnx::Conv_678 = Identity(%onnx::Conv_660)
%onnx::Conv_675 = Identity(%onnx::Conv_654)
%onnx::Conv_672 = Identity(%onnx::Conv_654)
%onnx::Conv_669 = Identity(%onnx::Conv_660)
%onnx::Conv_666 = Identity(%onnx::Conv_654)
%onnx::Conv_663 = Identity(%onnx::Conv_654)
%onnx::Conv_657 = Identity(%onnx::Conv_654)
%onnx::Conv_651 = Identity(%onnx::Conv_624)
%onnx::Conv_648 = Identity(%onnx::Conv_624)
%onnx::Conv_645 = Identity(%onnx::Conv_624)
%onnx::Conv_642 = Identity(%onnx::Conv_624)
%onnx::Conv_639 = Identity(%onnx::Conv_582)
%onnx::Conv_636 = Identity(%onnx::Conv_582)
%onnx::Conv_633 = Identity(%onnx::Conv_624)
%onnx::Conv_630 = Identity(%onnx::Conv_582)
%onnx::Conv_627 = Identity(%onnx::Conv_582)
%onnx::Conv_621 = Identity(%onnx::Conv_597)
%onnx::Conv_618 = Identity(%onnx::Conv_597)
%onnx::Conv_615 = Identity(%onnx::Conv_597)
%onnx::Conv_612 = Identity(%onnx::Conv_597)
%onnx::Conv_609 = Identity(%onnx::Conv_597)
%onnx::Conv_606 = Identity(%onnx::Conv_597)
%onnx::Conv_603 = Identity(%onnx::Conv_600)
%onnx::Conv_594 = Identity(%onnx::Conv_591)
%onnx::Conv_588 = Identity(%onnx::Conv_579)
%onnx::Conv_585 = Identity(%onnx::Conv_582)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_578, %onnx::Conv_579)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_581, %onnx::Conv_582)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_584, %onnx::Conv_585)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_587, %onnx::Conv_588)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_590, %onnx::Conv_591)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_593, %onnx::Conv_594)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_596, %onnx::Conv_597)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_599, %onnx::Conv_600)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_602, %onnx::Conv_603)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_605, %onnx::Conv_606)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_608, %onnx::Conv_609)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_611, %onnx::Conv_612)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_614, %onnx::Conv_615)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_617, %onnx::Conv_618)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_620, %onnx::Conv_621)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_623, %onnx::Conv_624)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%576 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %576
}
|
val_accuracy
| 0
| 74,628,352
| 1,630,796
|
{'zcp_synflow': 72.84295609254228, 'zcp_zen': 63.09949493408203, 'zcp_epe_nas': 27.92299439213907, 'zcp_fisher': 0.12444353103637695, 'zcp_flops': 74628352.0, 'zcp_grad_norm': 22.758756637573242, 'zcp_grasp': -0.07379341125488281, 'zcp_jacov': -16.054363917725745, 'zcp_l2_norm': 574.9521484375, 'zcp_nwot': 214.1109050507946, 'zcp_params': 1630796.0, 'zcp_plain': -0.003932628780603409, 'zcp_snip': 41.57341766357422, 'lat_1080ti_1': 0.40429261823900975, 'lat_1080ti_32': 0.45811827808572575, 'lat_1080ti_64': 0.46382363312297387, 'lat_2080ti_1': 0.39601748606762616, 'lat_2080ti_32': 0.42994404789563323, 'lat_2080ti_64': 0.45044813146942686, 'lat_essential_ph_1': 0.24528301886792453, 'lat_eyeriss': 0.493059248528414, 'lat_fpga': 0.5315014001721767, 'lat_gold_6226': 0.3253418417249103, 'lat_gold_6240': 0.3779355438826022, 'lat_pixel2': 0.41304347826086957, 'lat_pixel3': 0.5457740611171963, 'lat_raspi4': 0.5803158736758347, 'lat_samsung_a50': 0.18947368421052632, 'lat_samsung_s7': 0.1968503937007874, 'lat_silver_4114': 0.39874787757570773, 'lat_silver_4210r': 0.38699525452862693, 'lat_titan_rtx_1': 0.3855803857590262, 'lat_titan_rtx_32': 0.42812655412053235, 'lat_titan_rtx_64': 0.4559399097162637, 'lat_titanx_1': 0.20001802020919976, 'lat_titanx_32': 0.448461197168234, 'lat_titanx_64': 0.4535723475037443, 'lat_titanxp_1': 0.43804476494245476, 'lat_titanxp_32': 0.4515897039885308, 'lat_titanxp_64': 0.4830639659417529}
| |
FBNet_2698
|
FBNet
|
2698
|
2698
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_677[FLOAT, 16x3x3x3]
%onnx::Conv_678[FLOAT, 16]
%onnx::Conv_680[FLOAT, 16x8x1x1]
%onnx::Conv_683[FLOAT, 16x1x5x5]
%onnx::Conv_686[FLOAT, 24x8x1x1]
%onnx::Conv_687[FLOAT, 24]
%onnx::Conv_689[FLOAT, 72x24x1x1]
%onnx::Conv_690[FLOAT, 72]
%onnx::Conv_692[FLOAT, 72x1x5x5]
%onnx::Conv_695[FLOAT, 24x72x1x1]
%onnx::Conv_698[FLOAT, 24x24x1x1]
%onnx::Conv_701[FLOAT, 24x1x5x5]
%onnx::Conv_704[FLOAT, 24x24x1x1]
%onnx::Conv_707[FLOAT, 72x24x1x1]
%onnx::Conv_710[FLOAT, 72x1x3x3]
%onnx::Conv_713[FLOAT, 32x72x1x1]
%onnx::Conv_714[FLOAT, 32]
%onnx::Conv_716[FLOAT, 32x16x1x1]
%onnx::Conv_719[FLOAT, 32x1x3x3]
%onnx::Conv_722[FLOAT, 32x16x1x1]
%onnx::Conv_725[FLOAT, 96x32x1x1]
%onnx::Conv_726[FLOAT, 96]
%onnx::Conv_728[FLOAT, 96x1x3x3]
%onnx::Conv_731[FLOAT, 32x96x1x1]
%onnx::Conv_734[FLOAT, 96x32x1x1]
%onnx::Conv_737[FLOAT, 96x1x3x3]
%onnx::Conv_740[FLOAT, 32x96x1x1]
%onnx::Conv_743[FLOAT, 32x16x1x1]
%onnx::Conv_746[FLOAT, 32x1x3x3]
%onnx::Conv_749[FLOAT, 64x16x1x1]
%onnx::Conv_750[FLOAT, 64]
%onnx::Conv_752[FLOAT, 192x64x1x1]
%onnx::Conv_753[FLOAT, 192]
%onnx::Conv_755[FLOAT, 192x1x5x5]
%onnx::Conv_758[FLOAT, 64x192x1x1]
%onnx::Conv_761[FLOAT, 64x32x1x1]
%onnx::Conv_764[FLOAT, 64x1x3x3]
%onnx::Conv_767[FLOAT, 64x32x1x1]
%onnx::Conv_770[FLOAT, 192x64x1x1]
%onnx::Conv_773[FLOAT, 192x1x5x5]
%onnx::Conv_776[FLOAT, 64x192x1x1]
%onnx::Conv_779[FLOAT, 64x64x1x1]
%onnx::Conv_782[FLOAT, 64x1x3x3]
%onnx::Conv_785[FLOAT, 112x64x1x1]
%onnx::Conv_786[FLOAT, 112]
%onnx::Conv_788[FLOAT, 112x56x1x1]
%onnx::Conv_791[FLOAT, 112x1x3x3]
%onnx::Conv_794[FLOAT, 112x56x1x1]
%onnx::Conv_797[FLOAT, 336x112x1x1]
%onnx::Conv_798[FLOAT, 336]
%onnx::Conv_800[FLOAT, 336x1x5x5]
%onnx::Conv_803[FLOAT, 112x336x1x1]
%onnx::Conv_806[FLOAT, 112x56x1x1]
%onnx::Conv_809[FLOAT, 112x1x3x3]
%onnx::Conv_812[FLOAT, 112x56x1x1]
%onnx::Conv_815[FLOAT, 336x112x1x1]
%onnx::Conv_818[FLOAT, 336x1x5x5]
%onnx::Conv_821[FLOAT, 184x336x1x1]
%onnx::Conv_822[FLOAT, 184]
%onnx::Conv_824[FLOAT, 184x184x1x1]
%onnx::Conv_827[FLOAT, 184x1x5x5]
%onnx::Conv_830[FLOAT, 184x184x1x1]
%onnx::Conv_833[FLOAT, 1104x184x1x1]
%onnx::Conv_834[FLOAT, 1104]
%onnx::Conv_836[FLOAT, 1104x1x3x3]
%onnx::Conv_839[FLOAT, 184x1104x1x1]
%onnx::Conv_842[FLOAT, 1104x184x1x1]
%onnx::Conv_845[FLOAT, 1104x1x5x5]
%onnx::Conv_848[FLOAT, 184x1104x1x1]
%onnx::Conv_851[FLOAT, 552x184x1x1]
%onnx::Conv_852[FLOAT, 552]
%onnx::Conv_854[FLOAT, 552x1x3x3]
%onnx::Conv_857[FLOAT, 352x552x1x1]
%onnx::Conv_858[FLOAT, 352]
%onnx::Conv_860[FLOAT, 1504x352x1x1]
%onnx::Conv_861[FLOAT, 1504]
) {
%onnx::Conv_855 = Identity(%onnx::Conv_852)
%onnx::Conv_849 = Identity(%onnx::Conv_822)
%onnx::Conv_846 = Identity(%onnx::Conv_834)
%onnx::Conv_843 = Identity(%onnx::Conv_834)
%onnx::Conv_840 = Identity(%onnx::Conv_822)
%onnx::Conv_837 = Identity(%onnx::Conv_834)
%onnx::Conv_831 = Identity(%onnx::Conv_822)
%onnx::Conv_828 = Identity(%onnx::Conv_822)
%onnx::Conv_825 = Identity(%onnx::Conv_822)
%onnx::Conv_819 = Identity(%onnx::Conv_798)
%onnx::Conv_816 = Identity(%onnx::Conv_798)
%onnx::Conv_813 = Identity(%onnx::Conv_786)
%onnx::Conv_810 = Identity(%onnx::Conv_786)
%onnx::Conv_807 = Identity(%onnx::Conv_786)
%onnx::Conv_804 = Identity(%onnx::Conv_786)
%onnx::Conv_801 = Identity(%onnx::Conv_798)
%onnx::Conv_795 = Identity(%onnx::Conv_786)
%onnx::Conv_792 = Identity(%onnx::Conv_786)
%onnx::Conv_789 = Identity(%onnx::Conv_786)
%onnx::Conv_783 = Identity(%onnx::Conv_750)
%onnx::Conv_780 = Identity(%onnx::Conv_750)
%onnx::Conv_777 = Identity(%onnx::Conv_750)
%onnx::Conv_774 = Identity(%onnx::Conv_753)
%onnx::Conv_771 = Identity(%onnx::Conv_753)
%onnx::Conv_768 = Identity(%onnx::Conv_750)
%onnx::Conv_765 = Identity(%onnx::Conv_750)
%onnx::Conv_762 = Identity(%onnx::Conv_750)
%onnx::Conv_759 = Identity(%onnx::Conv_750)
%onnx::Conv_756 = Identity(%onnx::Conv_753)
%onnx::Conv_747 = Identity(%onnx::Conv_714)
%onnx::Conv_744 = Identity(%onnx::Conv_714)
%onnx::Conv_741 = Identity(%onnx::Conv_714)
%onnx::Conv_738 = Identity(%onnx::Conv_726)
%onnx::Conv_735 = Identity(%onnx::Conv_726)
%onnx::Conv_732 = Identity(%onnx::Conv_714)
%onnx::Conv_729 = Identity(%onnx::Conv_726)
%onnx::Conv_723 = Identity(%onnx::Conv_714)
%onnx::Conv_720 = Identity(%onnx::Conv_714)
%onnx::Conv_717 = Identity(%onnx::Conv_714)
%onnx::Conv_711 = Identity(%onnx::Conv_690)
%onnx::Conv_708 = Identity(%onnx::Conv_690)
%onnx::Conv_705 = Identity(%onnx::Conv_687)
%onnx::Conv_702 = Identity(%onnx::Conv_687)
%onnx::Conv_699 = Identity(%onnx::Conv_687)
%onnx::Conv_696 = Identity(%onnx::Conv_687)
%onnx::Conv_693 = Identity(%onnx::Conv_690)
%onnx::Conv_684 = Identity(%onnx::Conv_678)
%onnx::Conv_681 = Identity(%onnx::Conv_678)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_677, %onnx::Conv_678)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_803, %onnx::Conv_804)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_806, %onnx::Conv_807)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_809, %onnx::Conv_810)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_812, %onnx::Conv_813)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_815, %onnx::Conv_816)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_818, %onnx::Conv_819)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_821, %onnx::Conv_822)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_824, %onnx::Conv_825)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_827, %onnx::Conv_828)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_830, %onnx::Conv_831)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_833, %onnx::Conv_834)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_836, %onnx::Conv_837)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_839, %onnx::Conv_840)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_842, %onnx::Conv_843)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_845, %onnx::Conv_846)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_848, %onnx::Conv_849)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_851, %onnx::Conv_852)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_854, %onnx::Conv_855)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_857, %onnx::Conv_858)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_860, %onnx::Conv_861)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%675 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %675
}
|
val_accuracy
| 0
| 60,857,216
| 2,253,212
|
{'zcp_synflow': 74.66493833170333, 'zcp_zen': 67.9677734375, 'zcp_epe_nas': 16.277934268395352, 'zcp_fisher': 0.07068055868148804, 'zcp_flops': 60857216.0, 'zcp_grad_norm': 20.359176635742188, 'zcp_grasp': -0.05605125427246094, 'zcp_jacov': -16.058852519599647, 'zcp_l2_norm': 631.54736328125, 'zcp_nwot': 206.7301474808231, 'zcp_params': 2253212.0, 'zcp_plain': 0.0020908452570438385, 'zcp_snip': 36.985130310058594, 'lat_1080ti_1': 0.6168123944754245, 'lat_1080ti_32': 0.49703753445671, 'lat_1080ti_64': 0.2760442775683169, 'lat_2080ti_1': 0.6391216595833059, 'lat_2080ti_32': 0.5269421983524216, 'lat_2080ti_64': 0.3198778989081777, 'lat_essential_ph_1': 0.22641509433962265, 'lat_eyeriss': 0.34628573524288025, 'lat_fpga': 0.37892716136118504, 'lat_gold_6226': 0.38853397472547146, 'lat_gold_6240': 0.5250015003067096, 'lat_pixel2': 0.2826086956521739, 'lat_pixel3': 0.3397796112092286, 'lat_raspi4': 0.44027317227940066, 'lat_samsung_a50': 0.25263157894736843, 'lat_samsung_s7': 0.1732283464566929, 'lat_silver_4114': 0.64402593152879, 'lat_silver_4210r': 0.531330880624663, 'lat_titan_rtx_1': 0.6023941412173743, 'lat_titan_rtx_32': 0.5259430598764709, 'lat_titan_rtx_64': 0.35548922098407254, 'lat_titanx_1': 0.3346575726815163, 'lat_titanx_32': 0.4463631692790811, 'lat_titanx_64': 0.3157398916692932, 'lat_titanxp_1': 0.5867311232503357, 'lat_titanxp_32': 0.4856088281413111, 'lat_titanxp_64': 0.3260461442916164}
| |
FBNet_4045
|
FBNet
|
4045
|
4045
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_606[FLOAT, 16x3x3x3]
%onnx::Conv_607[FLOAT, 16]
%onnx::Conv_609[FLOAT, 16x8x1x1]
%onnx::Conv_612[FLOAT, 16x1x3x3]
%onnx::Conv_615[FLOAT, 16x8x1x1]
%onnx::Conv_618[FLOAT, 16x16x1x1]
%onnx::Conv_621[FLOAT, 16x1x3x3]
%onnx::Conv_624[FLOAT, 24x16x1x1]
%onnx::Conv_625[FLOAT, 24]
%onnx::Conv_627[FLOAT, 72x24x1x1]
%onnx::Conv_628[FLOAT, 72]
%onnx::Conv_630[FLOAT, 72x1x5x5]
%onnx::Conv_633[FLOAT, 24x72x1x1]
%onnx::Conv_636[FLOAT, 24x24x1x1]
%onnx::Conv_639[FLOAT, 24x1x3x3]
%onnx::Conv_642[FLOAT, 24x24x1x1]
%onnx::Conv_645[FLOAT, 24x24x1x1]
%onnx::Conv_648[FLOAT, 24x1x3x3]
%onnx::Conv_651[FLOAT, 24x24x1x1]
%onnx::Conv_654[FLOAT, 144x24x1x1]
%onnx::Conv_655[FLOAT, 144]
%onnx::Conv_657[FLOAT, 144x1x5x5]
%onnx::Conv_660[FLOAT, 32x144x1x1]
%onnx::Conv_661[FLOAT, 32]
%onnx::Conv_663[FLOAT, 96x32x1x1]
%onnx::Conv_664[FLOAT, 96]
%onnx::Conv_666[FLOAT, 96x1x3x3]
%onnx::Conv_669[FLOAT, 32x96x1x1]
%onnx::Conv_672[FLOAT, 192x32x1x1]
%onnx::Conv_673[FLOAT, 192]
%onnx::Conv_675[FLOAT, 192x1x5x5]
%onnx::Conv_678[FLOAT, 32x192x1x1]
%onnx::Conv_681[FLOAT, 96x32x1x1]
%onnx::Conv_684[FLOAT, 96x1x5x5]
%onnx::Conv_687[FLOAT, 64x96x1x1]
%onnx::Conv_688[FLOAT, 64]
%onnx::Conv_690[FLOAT, 192x64x1x1]
%onnx::Conv_693[FLOAT, 192x1x3x3]
%onnx::Conv_696[FLOAT, 64x192x1x1]
%onnx::Conv_699[FLOAT, 64x64x1x1]
%onnx::Conv_702[FLOAT, 64x1x5x5]
%onnx::Conv_705[FLOAT, 64x64x1x1]
%onnx::Conv_708[FLOAT, 384x64x1x1]
%onnx::Conv_709[FLOAT, 384]
%onnx::Conv_711[FLOAT, 384x1x3x3]
%onnx::Conv_714[FLOAT, 112x384x1x1]
%onnx::Conv_715[FLOAT, 112]
%onnx::Conv_717[FLOAT, 112x56x1x1]
%onnx::Conv_720[FLOAT, 112x1x5x5]
%onnx::Conv_723[FLOAT, 112x56x1x1]
%onnx::Conv_726[FLOAT, 336x112x1x1]
%onnx::Conv_727[FLOAT, 336]
%onnx::Conv_729[FLOAT, 336x1x5x5]
%onnx::Conv_732[FLOAT, 112x336x1x1]
%onnx::Conv_735[FLOAT, 112x56x1x1]
%onnx::Conv_738[FLOAT, 112x1x3x3]
%onnx::Conv_741[FLOAT, 112x56x1x1]
%onnx::Conv_744[FLOAT, 184x112x1x1]
%onnx::Conv_745[FLOAT, 184]
%onnx::Conv_747[FLOAT, 184x92x1x1]
%onnx::Conv_750[FLOAT, 184x1x3x3]
%onnx::Conv_753[FLOAT, 184x92x1x1]
%onnx::Conv_756[FLOAT, 552x184x1x1]
%onnx::Conv_757[FLOAT, 552]
%onnx::Conv_759[FLOAT, 552x1x5x5]
%onnx::Conv_762[FLOAT, 184x552x1x1]
%onnx::Conv_765[FLOAT, 552x184x1x1]
%onnx::Conv_768[FLOAT, 552x1x3x3]
%onnx::Conv_771[FLOAT, 184x552x1x1]
%onnx::Conv_774[FLOAT, 352x184x1x1]
%onnx::Conv_775[FLOAT, 352]
%onnx::Conv_777[FLOAT, 1504x352x1x1]
%onnx::Conv_778[FLOAT, 1504]
) {
%onnx::Conv_772 = Identity(%onnx::Conv_745)
%onnx::Conv_769 = Identity(%onnx::Conv_757)
%onnx::Conv_766 = Identity(%onnx::Conv_757)
%onnx::Conv_763 = Identity(%onnx::Conv_745)
%onnx::Conv_760 = Identity(%onnx::Conv_757)
%onnx::Conv_754 = Identity(%onnx::Conv_745)
%onnx::Conv_751 = Identity(%onnx::Conv_745)
%onnx::Conv_748 = Identity(%onnx::Conv_745)
%onnx::Conv_742 = Identity(%onnx::Conv_715)
%onnx::Conv_739 = Identity(%onnx::Conv_715)
%onnx::Conv_736 = Identity(%onnx::Conv_715)
%onnx::Conv_733 = Identity(%onnx::Conv_715)
%onnx::Conv_730 = Identity(%onnx::Conv_727)
%onnx::Conv_724 = Identity(%onnx::Conv_715)
%onnx::Conv_721 = Identity(%onnx::Conv_715)
%onnx::Conv_718 = Identity(%onnx::Conv_715)
%onnx::Conv_712 = Identity(%onnx::Conv_709)
%onnx::Conv_706 = Identity(%onnx::Conv_688)
%onnx::Conv_703 = Identity(%onnx::Conv_688)
%onnx::Conv_700 = Identity(%onnx::Conv_688)
%onnx::Conv_697 = Identity(%onnx::Conv_688)
%onnx::Conv_694 = Identity(%onnx::Conv_673)
%onnx::Conv_691 = Identity(%onnx::Conv_673)
%onnx::Conv_685 = Identity(%onnx::Conv_664)
%onnx::Conv_682 = Identity(%onnx::Conv_664)
%onnx::Conv_679 = Identity(%onnx::Conv_661)
%onnx::Conv_676 = Identity(%onnx::Conv_673)
%onnx::Conv_670 = Identity(%onnx::Conv_661)
%onnx::Conv_667 = Identity(%onnx::Conv_664)
%onnx::Conv_658 = Identity(%onnx::Conv_655)
%onnx::Conv_652 = Identity(%onnx::Conv_625)
%onnx::Conv_649 = Identity(%onnx::Conv_625)
%onnx::Conv_646 = Identity(%onnx::Conv_625)
%onnx::Conv_643 = Identity(%onnx::Conv_625)
%onnx::Conv_640 = Identity(%onnx::Conv_625)
%onnx::Conv_637 = Identity(%onnx::Conv_625)
%onnx::Conv_634 = Identity(%onnx::Conv_625)
%onnx::Conv_631 = Identity(%onnx::Conv_628)
%onnx::Conv_622 = Identity(%onnx::Conv_607)
%onnx::Conv_619 = Identity(%onnx::Conv_607)
%onnx::Conv_616 = Identity(%onnx::Conv_607)
%onnx::Conv_613 = Identity(%onnx::Conv_607)
%onnx::Conv_610 = Identity(%onnx::Conv_607)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_606, %onnx::Conv_607)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_609, %onnx::Conv_610)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_612, %onnx::Conv_613)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_615, %onnx::Conv_616)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_618, %onnx::Conv_619)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_621, %onnx::Conv_622)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_624, %onnx::Conv_625)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_627, %onnx::Conv_628)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_630, %onnx::Conv_631)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_output_0 = Reshape(%/cells.18/nl/Relu_output_0, %/cells.18/shuffle/Constant_output_0)
%/cells.18/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.18/shuffle/Reshape_output_0)
%/cells.18/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_1_output_0 = Reshape(%/cells.18/shuffle/Transpose_output_0, %/cells.18/shuffle/Constant_1_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/shuffle/Reshape_1_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.21/relu/Relu_output_0 = Relu(%/cells.21/conv/Conv_output_0)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/relu/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%604 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %604
}
|
val_accuracy
| 0
| 57,053,184
| 1,522,100
|
{'zcp_synflow': 74.11257311129339, 'zcp_zen': 62.341026306152344, 'zcp_epe_nas': 16.289505283999524, 'zcp_fisher': 0.08119790256023407, 'zcp_flops': 57053184.0, 'zcp_grad_norm': 18.998035430908203, 'zcp_grasp': -0.023969650268554688, 'zcp_jacov': -16.0530670054188, 'zcp_l2_norm': 557.0287475585938, 'zcp_nwot': 210.37885381896112, 'zcp_params': 1522100.0, 'zcp_plain': 0.004515163134783506, 'zcp_snip': 29.5582275390625, 'lat_1080ti_1': 0.4363832809902425, 'lat_1080ti_32': 0.43944437847912565, 'lat_1080ti_64': 0.3195518135923864, 'lat_2080ti_1': 0.48262925502610415, 'lat_2080ti_32': 0.4178313906335454, 'lat_2080ti_64': 0.3382445079202768, 'lat_essential_ph_1': 0.22641509433962265, 'lat_eyeriss': 0.2966919113197512, 'lat_fpga': 0.27780801314083836, 'lat_gold_6226': 0.25331372020753934, 'lat_gold_6240': 0.31471635145471166, 'lat_pixel2': 0.1956521739130435, 'lat_pixel3': 0.30802286759332886, 'lat_raspi4': 0.30163641812135145, 'lat_samsung_a50': 0.12631578947368421, 'lat_samsung_s7': 0.07086614173228346, 'lat_silver_4114': 0.33263223329339514, 'lat_silver_4210r': 0.3296768495286746, 'lat_titan_rtx_1': 0.45346612420768584, 'lat_titan_rtx_32': 0.4034427285452333, 'lat_titan_rtx_64': 0.3502883970445046, 'lat_titanx_1': 0.2346417414872565, 'lat_titanx_32': 0.3529868419875791, 'lat_titanx_64': 0.31850629565584565, 'lat_titanxp_1': 0.4150477857949668, 'lat_titanxp_32': 0.37504223115981444, 'lat_titanxp_64': 0.334307602022656}
| |
FBNet_4359
|
FBNet
|
4359
|
4359
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_678[FLOAT, 16x3x3x3]
%onnx::Conv_679[FLOAT, 16]
%onnx::Conv_681[FLOAT, 16x8x1x1]
%onnx::Conv_684[FLOAT, 16x1x3x3]
%onnx::Conv_687[FLOAT, 16x8x1x1]
%onnx::Conv_690[FLOAT, 24x16x1x1]
%onnx::Conv_691[FLOAT, 24]
%onnx::Conv_693[FLOAT, 24x12x1x1]
%onnx::Conv_696[FLOAT, 24x1x3x3]
%onnx::Conv_699[FLOAT, 24x12x1x1]
%onnx::Conv_702[FLOAT, 24x12x1x1]
%onnx::Conv_705[FLOAT, 24x1x5x5]
%onnx::Conv_708[FLOAT, 24x12x1x1]
%onnx::Conv_711[FLOAT, 24x12x1x1]
%onnx::Conv_714[FLOAT, 24x1x5x5]
%onnx::Conv_717[FLOAT, 32x12x1x1]
%onnx::Conv_718[FLOAT, 32]
%onnx::Conv_720[FLOAT, 192x32x1x1]
%onnx::Conv_721[FLOAT, 192]
%onnx::Conv_723[FLOAT, 192x1x3x3]
%onnx::Conv_726[FLOAT, 32x192x1x1]
%onnx::Conv_729[FLOAT, 192x32x1x1]
%onnx::Conv_732[FLOAT, 192x1x3x3]
%onnx::Conv_735[FLOAT, 32x192x1x1]
%onnx::Conv_738[FLOAT, 192x32x1x1]
%onnx::Conv_741[FLOAT, 192x1x5x5]
%onnx::Conv_744[FLOAT, 32x192x1x1]
%onnx::Conv_747[FLOAT, 192x32x1x1]
%onnx::Conv_750[FLOAT, 192x1x3x3]
%onnx::Conv_753[FLOAT, 64x192x1x1]
%onnx::Conv_754[FLOAT, 64]
%onnx::Conv_756[FLOAT, 64x32x1x1]
%onnx::Conv_759[FLOAT, 64x1x3x3]
%onnx::Conv_762[FLOAT, 64x32x1x1]
%onnx::Conv_765[FLOAT, 384x64x1x1]
%onnx::Conv_766[FLOAT, 384]
%onnx::Conv_768[FLOAT, 384x1x3x3]
%onnx::Conv_771[FLOAT, 64x384x1x1]
%onnx::Conv_774[FLOAT, 384x64x1x1]
%onnx::Conv_777[FLOAT, 384x1x5x5]
%onnx::Conv_780[FLOAT, 112x384x1x1]
%onnx::Conv_781[FLOAT, 112]
%onnx::Conv_783[FLOAT, 112x112x1x1]
%onnx::Conv_786[FLOAT, 112x1x3x3]
%onnx::Conv_789[FLOAT, 112x112x1x1]
%onnx::Conv_792[FLOAT, 336x112x1x1]
%onnx::Conv_793[FLOAT, 336]
%onnx::Conv_795[FLOAT, 336x1x3x3]
%onnx::Conv_798[FLOAT, 112x336x1x1]
%onnx::Conv_801[FLOAT, 672x112x1x1]
%onnx::Conv_802[FLOAT, 672]
%onnx::Conv_804[FLOAT, 672x1x5x5]
%onnx::Conv_807[FLOAT, 112x672x1x1]
%onnx::Conv_810[FLOAT, 336x112x1x1]
%onnx::Conv_813[FLOAT, 336x1x3x3]
%onnx::Conv_816[FLOAT, 184x336x1x1]
%onnx::Conv_817[FLOAT, 184]
%onnx::Conv_819[FLOAT, 184x92x1x1]
%onnx::Conv_822[FLOAT, 184x1x5x5]
%onnx::Conv_825[FLOAT, 184x92x1x1]
%onnx::Conv_828[FLOAT, 184x92x1x1]
%onnx::Conv_831[FLOAT, 184x1x3x3]
%onnx::Conv_834[FLOAT, 184x92x1x1]
%onnx::Conv_837[FLOAT, 552x184x1x1]
%onnx::Conv_838[FLOAT, 552]
%onnx::Conv_840[FLOAT, 552x1x3x3]
%onnx::Conv_843[FLOAT, 184x552x1x1]
%onnx::Conv_846[FLOAT, 552x184x1x1]
%onnx::Conv_849[FLOAT, 552x1x5x5]
%onnx::Conv_852[FLOAT, 352x552x1x1]
%onnx::Conv_853[FLOAT, 352]
%onnx::Conv_855[FLOAT, 1504x352x1x1]
%onnx::Conv_856[FLOAT, 1504]
) {
%onnx::Conv_850 = Identity(%onnx::Conv_838)
%onnx::Conv_847 = Identity(%onnx::Conv_838)
%onnx::Conv_844 = Identity(%onnx::Conv_817)
%onnx::Conv_841 = Identity(%onnx::Conv_838)
%onnx::Conv_835 = Identity(%onnx::Conv_817)
%onnx::Conv_832 = Identity(%onnx::Conv_817)
%onnx::Conv_829 = Identity(%onnx::Conv_817)
%onnx::Conv_826 = Identity(%onnx::Conv_817)
%onnx::Conv_823 = Identity(%onnx::Conv_817)
%onnx::Conv_820 = Identity(%onnx::Conv_817)
%onnx::Conv_814 = Identity(%onnx::Conv_793)
%onnx::Conv_811 = Identity(%onnx::Conv_793)
%onnx::Conv_808 = Identity(%onnx::Conv_781)
%onnx::Conv_805 = Identity(%onnx::Conv_802)
%onnx::Conv_799 = Identity(%onnx::Conv_781)
%onnx::Conv_796 = Identity(%onnx::Conv_793)
%onnx::Conv_790 = Identity(%onnx::Conv_781)
%onnx::Conv_787 = Identity(%onnx::Conv_781)
%onnx::Conv_784 = Identity(%onnx::Conv_781)
%onnx::Conv_778 = Identity(%onnx::Conv_766)
%onnx::Conv_775 = Identity(%onnx::Conv_766)
%onnx::Conv_772 = Identity(%onnx::Conv_754)
%onnx::Conv_769 = Identity(%onnx::Conv_766)
%onnx::Conv_763 = Identity(%onnx::Conv_754)
%onnx::Conv_760 = Identity(%onnx::Conv_754)
%onnx::Conv_757 = Identity(%onnx::Conv_754)
%onnx::Conv_751 = Identity(%onnx::Conv_721)
%onnx::Conv_748 = Identity(%onnx::Conv_721)
%onnx::Conv_745 = Identity(%onnx::Conv_718)
%onnx::Conv_742 = Identity(%onnx::Conv_721)
%onnx::Conv_739 = Identity(%onnx::Conv_721)
%onnx::Conv_736 = Identity(%onnx::Conv_718)
%onnx::Conv_733 = Identity(%onnx::Conv_721)
%onnx::Conv_730 = Identity(%onnx::Conv_721)
%onnx::Conv_727 = Identity(%onnx::Conv_718)
%onnx::Conv_724 = Identity(%onnx::Conv_721)
%onnx::Conv_715 = Identity(%onnx::Conv_691)
%onnx::Conv_712 = Identity(%onnx::Conv_691)
%onnx::Conv_709 = Identity(%onnx::Conv_691)
%onnx::Conv_706 = Identity(%onnx::Conv_691)
%onnx::Conv_703 = Identity(%onnx::Conv_691)
%onnx::Conv_700 = Identity(%onnx::Conv_691)
%onnx::Conv_697 = Identity(%onnx::Conv_691)
%onnx::Conv_694 = Identity(%onnx::Conv_691)
%onnx::Conv_688 = Identity(%onnx::Conv_679)
%onnx::Conv_685 = Identity(%onnx::Conv_679)
%onnx::Conv_682 = Identity(%onnx::Conv_679)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_678, %onnx::Conv_679)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.1/relu/Relu_output_0 = Relu(%/cells.1/conv/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/relu/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.1/relu/Relu_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_output_0 = Reshape(%/cells.18/nl/Relu_output_0, %/cells.18/shuffle/Constant_output_0)
%/cells.18/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.18/shuffle/Reshape_output_0)
%/cells.18/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_1_output_0 = Reshape(%/cells.18/shuffle/Transpose_output_0, %/cells.18/shuffle/Constant_1_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/shuffle/Reshape_1_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%676 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %676
}
|
val_accuracy
| 0
| 67,959,680
| 1,875,436
|
{'zcp_synflow': 70.6584410702742, 'zcp_zen': 66.20524597167969, 'zcp_epe_nas': 18.734937605001576, 'zcp_fisher': 0.061980269849300385, 'zcp_flops': 67959680.0, 'zcp_grad_norm': 18.100627899169922, 'zcp_grasp': 0.002899169921875, 'zcp_jacov': -16.052210313654538, 'zcp_l2_norm': 625.5097045898438, 'zcp_nwot': 210.09689279752507, 'zcp_params': 1875436.0, 'zcp_plain': 0.002030061325058341, 'zcp_snip': 35.895912170410156, 'lat_1080ti_1': 0.5693967742319427, 'lat_1080ti_32': 0.43404931253952167, 'lat_1080ti_64': 0.26777664871286794, 'lat_2080ti_1': 0.6149736317904875, 'lat_2080ti_32': 0.5470962417848446, 'lat_2080ti_64': 0.33317777473990384, 'lat_essential_ph_1': 0.32075471698113206, 'lat_eyeriss': 0.39979461977151437, 'lat_fpga': 0.46346193807577746, 'lat_gold_6226': 0.5593933796484151, 'lat_gold_6240': 0.5782761969724528, 'lat_pixel2': 0.2826086956521739, 'lat_pixel3': 0.4096164293728135, 'lat_raspi4': 0.40813390528477383, 'lat_samsung_a50': 0.2, 'lat_samsung_s7': 0.1732283464566929, 'lat_silver_4114': 0.6089606392529613, 'lat_silver_4210r': 0.6510089506669755, 'lat_titan_rtx_1': 0.5677255803027965, 'lat_titan_rtx_32': 0.46127876104113336, 'lat_titan_rtx_64': 0.34693474585119644, 'lat_titanx_1': 0.2999440092017397, 'lat_titanx_32': 0.34588344566422213, 'lat_titanx_64': 0.31730830589062925, 'lat_titanxp_1': 0.528393062210487, 'lat_titanxp_32': 0.40063418944102774, 'lat_titanxp_64': 0.2887786771884996}
| |
FBNet_731
|
FBNet
|
731
|
731
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_713[FLOAT, 16x3x3x3]
%onnx::Conv_714[FLOAT, 16]
%onnx::Conv_716[FLOAT, 48x16x1x1]
%onnx::Conv_717[FLOAT, 48]
%onnx::Conv_719[FLOAT, 48x1x5x5]
%onnx::Conv_722[FLOAT, 16x48x1x1]
%onnx::Conv_725[FLOAT, 16x16x1x1]
%onnx::Conv_728[FLOAT, 16x1x5x5]
%onnx::Conv_731[FLOAT, 24x16x1x1]
%onnx::Conv_732[FLOAT, 24]
%onnx::Conv_734[FLOAT, 24x24x1x1]
%onnx::Conv_737[FLOAT, 24x1x3x3]
%onnx::Conv_740[FLOAT, 24x24x1x1]
%onnx::Conv_743[FLOAT, 72x24x1x1]
%onnx::Conv_744[FLOAT, 72]
%onnx::Conv_746[FLOAT, 72x1x5x5]
%onnx::Conv_749[FLOAT, 24x72x1x1]
%onnx::Conv_752[FLOAT, 72x24x1x1]
%onnx::Conv_755[FLOAT, 72x1x5x5]
%onnx::Conv_758[FLOAT, 24x72x1x1]
%onnx::Conv_761[FLOAT, 24x12x1x1]
%onnx::Conv_764[FLOAT, 24x1x5x5]
%onnx::Conv_767[FLOAT, 32x12x1x1]
%onnx::Conv_768[FLOAT, 32]
%onnx::Conv_770[FLOAT, 192x32x1x1]
%onnx::Conv_771[FLOAT, 192]
%onnx::Conv_773[FLOAT, 192x1x5x5]
%onnx::Conv_776[FLOAT, 32x192x1x1]
%onnx::Conv_779[FLOAT, 32x32x1x1]
%onnx::Conv_782[FLOAT, 32x1x5x5]
%onnx::Conv_785[FLOAT, 32x32x1x1]
%onnx::Conv_788[FLOAT, 192x32x1x1]
%onnx::Conv_791[FLOAT, 192x1x5x5]
%onnx::Conv_794[FLOAT, 32x192x1x1]
%onnx::Conv_797[FLOAT, 32x32x1x1]
%onnx::Conv_800[FLOAT, 32x1x5x5]
%onnx::Conv_803[FLOAT, 64x32x1x1]
%onnx::Conv_804[FLOAT, 64]
%onnx::Conv_806[FLOAT, 384x64x1x1]
%onnx::Conv_807[FLOAT, 384]
%onnx::Conv_809[FLOAT, 384x1x5x5]
%onnx::Conv_812[FLOAT, 64x384x1x1]
%onnx::Conv_815[FLOAT, 64x32x1x1]
%onnx::Conv_818[FLOAT, 64x1x3x3]
%onnx::Conv_821[FLOAT, 64x32x1x1]
%onnx::Conv_824[FLOAT, 64x32x1x1]
%onnx::Conv_827[FLOAT, 64x1x3x3]
%onnx::Conv_830[FLOAT, 64x32x1x1]
%onnx::Conv_833[FLOAT, 384x64x1x1]
%onnx::Conv_836[FLOAT, 384x1x5x5]
%onnx::Conv_839[FLOAT, 112x384x1x1]
%onnx::Conv_840[FLOAT, 112]
%onnx::Conv_842[FLOAT, 112x112x1x1]
%onnx::Conv_845[FLOAT, 112x1x3x3]
%onnx::Conv_848[FLOAT, 112x112x1x1]
%onnx::Conv_851[FLOAT, 336x112x1x1]
%onnx::Conv_852[FLOAT, 336]
%onnx::Conv_854[FLOAT, 336x1x3x3]
%onnx::Conv_857[FLOAT, 112x336x1x1]
%onnx::Conv_860[FLOAT, 672x112x1x1]
%onnx::Conv_861[FLOAT, 672]
%onnx::Conv_863[FLOAT, 672x1x5x5]
%onnx::Conv_866[FLOAT, 112x672x1x1]
%onnx::Conv_869[FLOAT, 112x56x1x1]
%onnx::Conv_872[FLOAT, 112x1x3x3]
%onnx::Conv_875[FLOAT, 184x56x1x1]
%onnx::Conv_876[FLOAT, 184]
%onnx::Conv_878[FLOAT, 1104x184x1x1]
%onnx::Conv_879[FLOAT, 1104]
%onnx::Conv_881[FLOAT, 1104x1x3x3]
%onnx::Conv_884[FLOAT, 184x1104x1x1]
%onnx::Conv_887[FLOAT, 1104x184x1x1]
%onnx::Conv_890[FLOAT, 1104x1x3x3]
%onnx::Conv_893[FLOAT, 184x1104x1x1]
%onnx::Conv_896[FLOAT, 184x92x1x1]
%onnx::Conv_899[FLOAT, 184x1x5x5]
%onnx::Conv_902[FLOAT, 184x92x1x1]
%onnx::Conv_905[FLOAT, 184x184x1x1]
%onnx::Conv_908[FLOAT, 184x1x5x5]
%onnx::Conv_911[FLOAT, 352x184x1x1]
%onnx::Conv_912[FLOAT, 352]
%onnx::Conv_914[FLOAT, 1504x352x1x1]
%onnx::Conv_915[FLOAT, 1504]
) {
%onnx::Conv_909 = Identity(%onnx::Conv_876)
%onnx::Conv_906 = Identity(%onnx::Conv_876)
%onnx::Conv_903 = Identity(%onnx::Conv_876)
%onnx::Conv_900 = Identity(%onnx::Conv_876)
%onnx::Conv_897 = Identity(%onnx::Conv_876)
%onnx::Conv_894 = Identity(%onnx::Conv_876)
%onnx::Conv_891 = Identity(%onnx::Conv_879)
%onnx::Conv_888 = Identity(%onnx::Conv_879)
%onnx::Conv_885 = Identity(%onnx::Conv_876)
%onnx::Conv_882 = Identity(%onnx::Conv_879)
%onnx::Conv_873 = Identity(%onnx::Conv_840)
%onnx::Conv_870 = Identity(%onnx::Conv_840)
%onnx::Conv_867 = Identity(%onnx::Conv_840)
%onnx::Conv_864 = Identity(%onnx::Conv_861)
%onnx::Conv_858 = Identity(%onnx::Conv_840)
%onnx::Conv_855 = Identity(%onnx::Conv_852)
%onnx::Conv_849 = Identity(%onnx::Conv_840)
%onnx::Conv_846 = Identity(%onnx::Conv_840)
%onnx::Conv_843 = Identity(%onnx::Conv_840)
%onnx::Conv_837 = Identity(%onnx::Conv_807)
%onnx::Conv_834 = Identity(%onnx::Conv_807)
%onnx::Conv_831 = Identity(%onnx::Conv_804)
%onnx::Conv_828 = Identity(%onnx::Conv_804)
%onnx::Conv_825 = Identity(%onnx::Conv_804)
%onnx::Conv_822 = Identity(%onnx::Conv_804)
%onnx::Conv_819 = Identity(%onnx::Conv_804)
%onnx::Conv_816 = Identity(%onnx::Conv_804)
%onnx::Conv_813 = Identity(%onnx::Conv_804)
%onnx::Conv_810 = Identity(%onnx::Conv_807)
%onnx::Conv_801 = Identity(%onnx::Conv_768)
%onnx::Conv_798 = Identity(%onnx::Conv_768)
%onnx::Conv_795 = Identity(%onnx::Conv_768)
%onnx::Conv_792 = Identity(%onnx::Conv_771)
%onnx::Conv_789 = Identity(%onnx::Conv_771)
%onnx::Conv_786 = Identity(%onnx::Conv_768)
%onnx::Conv_783 = Identity(%onnx::Conv_768)
%onnx::Conv_780 = Identity(%onnx::Conv_768)
%onnx::Conv_777 = Identity(%onnx::Conv_768)
%onnx::Conv_774 = Identity(%onnx::Conv_771)
%onnx::Conv_765 = Identity(%onnx::Conv_732)
%onnx::Conv_762 = Identity(%onnx::Conv_732)
%onnx::Conv_759 = Identity(%onnx::Conv_732)
%onnx::Conv_756 = Identity(%onnx::Conv_744)
%onnx::Conv_753 = Identity(%onnx::Conv_744)
%onnx::Conv_750 = Identity(%onnx::Conv_732)
%onnx::Conv_747 = Identity(%onnx::Conv_744)
%onnx::Conv_741 = Identity(%onnx::Conv_732)
%onnx::Conv_738 = Identity(%onnx::Conv_732)
%onnx::Conv_735 = Identity(%onnx::Conv_732)
%onnx::Conv_729 = Identity(%onnx::Conv_714)
%onnx::Conv_726 = Identity(%onnx::Conv_714)
%onnx::Conv_723 = Identity(%onnx::Conv_714)
%onnx::Conv_720 = Identity(%onnx::Conv_717)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_713, %onnx::Conv_714)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_803, %onnx::Conv_804)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_806, %onnx::Conv_807)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_809, %onnx::Conv_810)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_812, %onnx::Conv_813)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_815, %onnx::Conv_816)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_818, %onnx::Conv_819)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_821, %onnx::Conv_822)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_824, %onnx::Conv_825)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_827, %onnx::Conv_828)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_830, %onnx::Conv_831)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_833, %onnx::Conv_834)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_836, %onnx::Conv_837)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_839, %onnx::Conv_840)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_842, %onnx::Conv_843)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_845, %onnx::Conv_846)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_848, %onnx::Conv_849)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_851, %onnx::Conv_852)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_854, %onnx::Conv_855)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_857, %onnx::Conv_858)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_860, %onnx::Conv_861)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_863, %onnx::Conv_864)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_866, %onnx::Conv_867)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_869, %onnx::Conv_870)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_872, %onnx::Conv_873)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_875, %onnx::Conv_876)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_878, %onnx::Conv_879)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_881, %onnx::Conv_882)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_884, %onnx::Conv_885)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_887, %onnx::Conv_888)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_890, %onnx::Conv_891)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_893, %onnx::Conv_894)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_896, %onnx::Conv_897)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_899, %onnx::Conv_900)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_902, %onnx::Conv_903)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_905, %onnx::Conv_906)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_908, %onnx::Conv_909)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_911, %onnx::Conv_912)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_914, %onnx::Conv_915)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%711 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %711
}
|
val_accuracy
| 0
| 81,666,432
| 2,175,572
|
{'zcp_synflow': 84.50173068755723, 'zcp_zen': 74.63818359375, 'zcp_epe_nas': 26.007820346598226, 'zcp_fisher': 0.15402711927890778, 'zcp_flops': 81666432.0, 'zcp_grad_norm': 32.849449157714844, 'zcp_grasp': -0.18608474731445312, 'zcp_jacov': -16.050210707935797, 'zcp_l2_norm': 687.12548828125, 'zcp_nwot': 213.39370852853318, 'zcp_params': 2175572.0, 'zcp_plain': -0.0016209024470299482, 'zcp_snip': 57.604827880859375, 'lat_1080ti_1': 0.7863123918043547, 'lat_1080ti_32': 0.69668329805721, 'lat_1080ti_64': 0.6234485680743788, 'lat_2080ti_1': 0.8051630044539372, 'lat_2080ti_32': 0.6585813223386193, 'lat_2080ti_64': 0.5837632118211032, 'lat_essential_ph_1': 0.5849056603773585, 'lat_eyeriss': 0.6198265270570114, 'lat_fpga': 0.6471862971135672, 'lat_gold_6226': 0.4808511425749864, 'lat_gold_6240': 0.826418239809675, 'lat_pixel2': 0.41304347826086957, 'lat_pixel3': 0.633204038152066, 'lat_raspi4': 0.5986046745463506, 'lat_samsung_a50': 0.2631578947368421, 'lat_samsung_s7': 0.25196850393700787, 'lat_silver_4114': 0.7447272511592918, 'lat_silver_4210r': 0.7888612016770691, 'lat_titan_rtx_1': 0.7759639222167692, 'lat_titan_rtx_32': 0.6594186118719212, 'lat_titan_rtx_64': 0.6382478935927225, 'lat_titanx_1': 0.41820104770354166, 'lat_titanx_32': 0.6625826812187612, 'lat_titanx_64': 0.5809839373679327, 'lat_titanxp_1': 0.7305793316815122, 'lat_titanxp_32': 0.6676190798031204, 'lat_titanxp_64': 0.6133419015563293}
| |
FBNet_4838
|
FBNet
|
4838
|
4838
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_687[FLOAT, 16x3x3x3]
%onnx::Conv_688[FLOAT, 16]
%onnx::Conv_690[FLOAT, 48x16x1x1]
%onnx::Conv_691[FLOAT, 48]
%onnx::Conv_693[FLOAT, 48x1x5x5]
%onnx::Conv_696[FLOAT, 16x48x1x1]
%onnx::Conv_699[FLOAT, 16x8x1x1]
%onnx::Conv_702[FLOAT, 16x1x5x5]
%onnx::Conv_705[FLOAT, 24x8x1x1]
%onnx::Conv_706[FLOAT, 24]
%onnx::Conv_708[FLOAT, 24x12x1x1]
%onnx::Conv_711[FLOAT, 24x1x3x3]
%onnx::Conv_714[FLOAT, 24x12x1x1]
%onnx::Conv_717[FLOAT, 72x24x1x1]
%onnx::Conv_718[FLOAT, 72]
%onnx::Conv_720[FLOAT, 72x1x3x3]
%onnx::Conv_723[FLOAT, 24x72x1x1]
%onnx::Conv_726[FLOAT, 72x24x1x1]
%onnx::Conv_729[FLOAT, 72x1x3x3]
%onnx::Conv_732[FLOAT, 32x72x1x1]
%onnx::Conv_733[FLOAT, 32]
%onnx::Conv_735[FLOAT, 32x16x1x1]
%onnx::Conv_738[FLOAT, 32x1x3x3]
%onnx::Conv_741[FLOAT, 32x16x1x1]
%onnx::Conv_744[FLOAT, 32x32x1x1]
%onnx::Conv_747[FLOAT, 32x1x3x3]
%onnx::Conv_750[FLOAT, 32x32x1x1]
%onnx::Conv_753[FLOAT, 96x32x1x1]
%onnx::Conv_754[FLOAT, 96]
%onnx::Conv_756[FLOAT, 96x1x3x3]
%onnx::Conv_759[FLOAT, 32x96x1x1]
%onnx::Conv_762[FLOAT, 32x16x1x1]
%onnx::Conv_765[FLOAT, 32x1x3x3]
%onnx::Conv_768[FLOAT, 64x16x1x1]
%onnx::Conv_769[FLOAT, 64]
%onnx::Conv_771[FLOAT, 64x64x1x1]
%onnx::Conv_774[FLOAT, 64x1x3x3]
%onnx::Conv_777[FLOAT, 64x64x1x1]
%onnx::Conv_780[FLOAT, 384x64x1x1]
%onnx::Conv_781[FLOAT, 384]
%onnx::Conv_783[FLOAT, 384x1x3x3]
%onnx::Conv_786[FLOAT, 64x384x1x1]
%onnx::Conv_789[FLOAT, 192x64x1x1]
%onnx::Conv_790[FLOAT, 192]
%onnx::Conv_792[FLOAT, 192x1x5x5]
%onnx::Conv_795[FLOAT, 64x192x1x1]
%onnx::Conv_798[FLOAT, 64x64x1x1]
%onnx::Conv_801[FLOAT, 64x1x3x3]
%onnx::Conv_804[FLOAT, 112x64x1x1]
%onnx::Conv_805[FLOAT, 112]
%onnx::Conv_807[FLOAT, 672x112x1x1]
%onnx::Conv_808[FLOAT, 672]
%onnx::Conv_810[FLOAT, 672x1x3x3]
%onnx::Conv_813[FLOAT, 112x672x1x1]
%onnx::Conv_816[FLOAT, 112x112x1x1]
%onnx::Conv_819[FLOAT, 112x1x5x5]
%onnx::Conv_822[FLOAT, 112x112x1x1]
%onnx::Conv_825[FLOAT, 672x112x1x1]
%onnx::Conv_828[FLOAT, 672x1x3x3]
%onnx::Conv_831[FLOAT, 112x672x1x1]
%onnx::Conv_834[FLOAT, 184x112x1x1]
%onnx::Conv_835[FLOAT, 184]
%onnx::Conv_837[FLOAT, 184x184x1x1]
%onnx::Conv_840[FLOAT, 184x1x3x3]
%onnx::Conv_843[FLOAT, 184x184x1x1]
%onnx::Conv_846[FLOAT, 184x92x1x1]
%onnx::Conv_849[FLOAT, 184x1x3x3]
%onnx::Conv_852[FLOAT, 184x92x1x1]
%onnx::Conv_855[FLOAT, 184x92x1x1]
%onnx::Conv_858[FLOAT, 184x1x3x3]
%onnx::Conv_861[FLOAT, 184x92x1x1]
%onnx::Conv_864[FLOAT, 1104x184x1x1]
%onnx::Conv_865[FLOAT, 1104]
%onnx::Conv_867[FLOAT, 1104x1x5x5]
%onnx::Conv_870[FLOAT, 352x1104x1x1]
%onnx::Conv_871[FLOAT, 352]
%onnx::Conv_873[FLOAT, 1504x352x1x1]
%onnx::Conv_874[FLOAT, 1504]
) {
%onnx::Conv_868 = Identity(%onnx::Conv_865)
%onnx::Conv_862 = Identity(%onnx::Conv_835)
%onnx::Conv_859 = Identity(%onnx::Conv_835)
%onnx::Conv_856 = Identity(%onnx::Conv_835)
%onnx::Conv_853 = Identity(%onnx::Conv_835)
%onnx::Conv_850 = Identity(%onnx::Conv_835)
%onnx::Conv_847 = Identity(%onnx::Conv_835)
%onnx::Conv_844 = Identity(%onnx::Conv_835)
%onnx::Conv_841 = Identity(%onnx::Conv_835)
%onnx::Conv_838 = Identity(%onnx::Conv_835)
%onnx::Conv_832 = Identity(%onnx::Conv_805)
%onnx::Conv_829 = Identity(%onnx::Conv_808)
%onnx::Conv_826 = Identity(%onnx::Conv_808)
%onnx::Conv_823 = Identity(%onnx::Conv_805)
%onnx::Conv_820 = Identity(%onnx::Conv_805)
%onnx::Conv_817 = Identity(%onnx::Conv_805)
%onnx::Conv_814 = Identity(%onnx::Conv_805)
%onnx::Conv_811 = Identity(%onnx::Conv_808)
%onnx::Conv_802 = Identity(%onnx::Conv_769)
%onnx::Conv_799 = Identity(%onnx::Conv_769)
%onnx::Conv_796 = Identity(%onnx::Conv_769)
%onnx::Conv_793 = Identity(%onnx::Conv_790)
%onnx::Conv_787 = Identity(%onnx::Conv_769)
%onnx::Conv_784 = Identity(%onnx::Conv_781)
%onnx::Conv_778 = Identity(%onnx::Conv_769)
%onnx::Conv_775 = Identity(%onnx::Conv_769)
%onnx::Conv_772 = Identity(%onnx::Conv_769)
%onnx::Conv_766 = Identity(%onnx::Conv_733)
%onnx::Conv_763 = Identity(%onnx::Conv_733)
%onnx::Conv_760 = Identity(%onnx::Conv_733)
%onnx::Conv_757 = Identity(%onnx::Conv_754)
%onnx::Conv_751 = Identity(%onnx::Conv_733)
%onnx::Conv_748 = Identity(%onnx::Conv_733)
%onnx::Conv_745 = Identity(%onnx::Conv_733)
%onnx::Conv_742 = Identity(%onnx::Conv_733)
%onnx::Conv_739 = Identity(%onnx::Conv_733)
%onnx::Conv_736 = Identity(%onnx::Conv_733)
%onnx::Conv_730 = Identity(%onnx::Conv_718)
%onnx::Conv_727 = Identity(%onnx::Conv_718)
%onnx::Conv_724 = Identity(%onnx::Conv_706)
%onnx::Conv_721 = Identity(%onnx::Conv_718)
%onnx::Conv_715 = Identity(%onnx::Conv_706)
%onnx::Conv_712 = Identity(%onnx::Conv_706)
%onnx::Conv_709 = Identity(%onnx::Conv_706)
%onnx::Conv_703 = Identity(%onnx::Conv_688)
%onnx::Conv_700 = Identity(%onnx::Conv_688)
%onnx::Conv_697 = Identity(%onnx::Conv_688)
%onnx::Conv_694 = Identity(%onnx::Conv_691)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_687, %onnx::Conv_688)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_858, %onnx::Conv_859)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_861, %onnx::Conv_862)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_864, %onnx::Conv_865)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_867, %onnx::Conv_868)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_870, %onnx::Conv_871)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_873, %onnx::Conv_874)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%685 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %685
}
|
val_accuracy
| 0
| 66,553,856
| 1,953,844
|
{'zcp_synflow': 75.1226817671799, 'zcp_zen': 66.04403686523438, 'zcp_epe_nas': 6.666528550379195, 'zcp_fisher': 0.12139669805765152, 'zcp_flops': 66553856.0, 'zcp_grad_norm': 26.297117233276367, 'zcp_grasp': 0.07900619506835938, 'zcp_jacov': -16.052885758767083, 'zcp_l2_norm': 611.0889892578125, 'zcp_nwot': 209.6788244856222, 'zcp_params': 1953844.0, 'zcp_plain': 0.003740076208487153, 'zcp_snip': 42.17448043823242, 'lat_1080ti_1': 0.7420529557438568, 'lat_1080ti_32': 0.4869619816941561, 'lat_1080ti_64': 0.3477395187734028, 'lat_2080ti_1': 0.7103719828801355, 'lat_2080ti_32': 0.5358433155031973, 'lat_2080ti_64': 0.3937452453825687, 'lat_essential_ph_1': 0.16981132075471697, 'lat_eyeriss': 0.36117580180807946, 'lat_fpga': 0.490218585182913, 'lat_gold_6226': 0.3312159045810847, 'lat_gold_6240': 0.44249731021473837, 'lat_pixel2': 0.2826086956521739, 'lat_pixel3': 0.3609918203251304, 'lat_raspi4': 0.4287049548580669, 'lat_samsung_a50': 0.16842105263157894, 'lat_samsung_s7': 0.16535433070866143, 'lat_silver_4114': 0.4836547251697923, 'lat_silver_4210r': 0.4768972084317991, 'lat_titan_rtx_1': 0.6514077795354836, 'lat_titan_rtx_32': 0.5379736844646156, 'lat_titan_rtx_64': 0.4078822322143383, 'lat_titanx_1': 0.34499710244704346, 'lat_titanx_32': 0.4421636880389099, 'lat_titanx_64': 0.3849437887050708, 'lat_titanxp_1': 0.6228599346589718, 'lat_titanxp_32': 0.49148272341487365, 'lat_titanxp_64': 0.36822419789944555}
| |
FBNet_3340
|
FBNet
|
3340
|
3340
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_570[FLOAT, 16x3x3x3]
%onnx::Conv_571[FLOAT, 16]
%onnx::Conv_573[FLOAT, 48x16x1x1]
%onnx::Conv_574[FLOAT, 48]
%onnx::Conv_576[FLOAT, 48x1x3x3]
%onnx::Conv_579[FLOAT, 16x48x1x1]
%onnx::Conv_582[FLOAT, 48x16x1x1]
%onnx::Conv_585[FLOAT, 48x1x3x3]
%onnx::Conv_588[FLOAT, 24x48x1x1]
%onnx::Conv_589[FLOAT, 24]
%onnx::Conv_591[FLOAT, 144x24x1x1]
%onnx::Conv_592[FLOAT, 144]
%onnx::Conv_594[FLOAT, 144x1x3x3]
%onnx::Conv_597[FLOAT, 24x144x1x1]
%onnx::Conv_600[FLOAT, 24x24x1x1]
%onnx::Conv_603[FLOAT, 24x1x3x3]
%onnx::Conv_606[FLOAT, 24x24x1x1]
%onnx::Conv_609[FLOAT, 32x24x1x1]
%onnx::Conv_610[FLOAT, 32]
%onnx::Conv_612[FLOAT, 96x32x1x1]
%onnx::Conv_613[FLOAT, 96]
%onnx::Conv_615[FLOAT, 96x1x3x3]
%onnx::Conv_618[FLOAT, 32x96x1x1]
%onnx::Conv_621[FLOAT, 32x16x1x1]
%onnx::Conv_624[FLOAT, 32x1x3x3]
%onnx::Conv_627[FLOAT, 32x16x1x1]
%onnx::Conv_630[FLOAT, 32x16x1x1]
%onnx::Conv_633[FLOAT, 32x1x3x3]
%onnx::Conv_636[FLOAT, 64x16x1x1]
%onnx::Conv_637[FLOAT, 64]
%onnx::Conv_639[FLOAT, 64x64x1x1]
%onnx::Conv_642[FLOAT, 64x1x5x5]
%onnx::Conv_645[FLOAT, 64x64x1x1]
%onnx::Conv_648[FLOAT, 384x64x1x1]
%onnx::Conv_649[FLOAT, 384]
%onnx::Conv_651[FLOAT, 384x1x3x3]
%onnx::Conv_654[FLOAT, 64x384x1x1]
%onnx::Conv_657[FLOAT, 64x64x1x1]
%onnx::Conv_660[FLOAT, 64x1x3x3]
%onnx::Conv_663[FLOAT, 64x64x1x1]
%onnx::Conv_666[FLOAT, 64x32x1x1]
%onnx::Conv_669[FLOAT, 64x1x5x5]
%onnx::Conv_672[FLOAT, 112x32x1x1]
%onnx::Conv_673[FLOAT, 112]
%onnx::Conv_675[FLOAT, 112x112x1x1]
%onnx::Conv_678[FLOAT, 112x1x5x5]
%onnx::Conv_681[FLOAT, 112x112x1x1]
%onnx::Conv_684[FLOAT, 112x112x1x1]
%onnx::Conv_687[FLOAT, 112x1x5x5]
%onnx::Conv_690[FLOAT, 112x112x1x1]
%onnx::Conv_693[FLOAT, 184x112x1x1]
%onnx::Conv_694[FLOAT, 184]
%onnx::Conv_696[FLOAT, 1104x184x1x1]
%onnx::Conv_697[FLOAT, 1104]
%onnx::Conv_699[FLOAT, 1104x1x3x3]
%onnx::Conv_702[FLOAT, 184x1104x1x1]
%onnx::Conv_705[FLOAT, 184x92x1x1]
%onnx::Conv_708[FLOAT, 184x1x5x5]
%onnx::Conv_711[FLOAT, 184x92x1x1]
%onnx::Conv_714[FLOAT, 184x92x1x1]
%onnx::Conv_717[FLOAT, 184x1x5x5]
%onnx::Conv_720[FLOAT, 352x92x1x1]
%onnx::Conv_721[FLOAT, 352]
%onnx::Conv_723[FLOAT, 1504x352x1x1]
%onnx::Conv_724[FLOAT, 1504]
) {
%onnx::Conv_718 = Identity(%onnx::Conv_694)
%onnx::Conv_715 = Identity(%onnx::Conv_694)
%onnx::Conv_712 = Identity(%onnx::Conv_694)
%onnx::Conv_709 = Identity(%onnx::Conv_694)
%onnx::Conv_706 = Identity(%onnx::Conv_694)
%onnx::Conv_703 = Identity(%onnx::Conv_694)
%onnx::Conv_700 = Identity(%onnx::Conv_697)
%onnx::Conv_691 = Identity(%onnx::Conv_673)
%onnx::Conv_688 = Identity(%onnx::Conv_673)
%onnx::Conv_685 = Identity(%onnx::Conv_673)
%onnx::Conv_682 = Identity(%onnx::Conv_673)
%onnx::Conv_679 = Identity(%onnx::Conv_673)
%onnx::Conv_676 = Identity(%onnx::Conv_673)
%onnx::Conv_670 = Identity(%onnx::Conv_637)
%onnx::Conv_667 = Identity(%onnx::Conv_637)
%onnx::Conv_664 = Identity(%onnx::Conv_637)
%onnx::Conv_661 = Identity(%onnx::Conv_637)
%onnx::Conv_658 = Identity(%onnx::Conv_637)
%onnx::Conv_655 = Identity(%onnx::Conv_637)
%onnx::Conv_652 = Identity(%onnx::Conv_649)
%onnx::Conv_646 = Identity(%onnx::Conv_637)
%onnx::Conv_643 = Identity(%onnx::Conv_637)
%onnx::Conv_640 = Identity(%onnx::Conv_637)
%onnx::Conv_634 = Identity(%onnx::Conv_610)
%onnx::Conv_631 = Identity(%onnx::Conv_610)
%onnx::Conv_628 = Identity(%onnx::Conv_610)
%onnx::Conv_625 = Identity(%onnx::Conv_610)
%onnx::Conv_622 = Identity(%onnx::Conv_610)
%onnx::Conv_619 = Identity(%onnx::Conv_610)
%onnx::Conv_616 = Identity(%onnx::Conv_613)
%onnx::Conv_607 = Identity(%onnx::Conv_589)
%onnx::Conv_604 = Identity(%onnx::Conv_589)
%onnx::Conv_601 = Identity(%onnx::Conv_589)
%onnx::Conv_598 = Identity(%onnx::Conv_589)
%onnx::Conv_595 = Identity(%onnx::Conv_592)
%onnx::Conv_586 = Identity(%onnx::Conv_574)
%onnx::Conv_583 = Identity(%onnx::Conv_574)
%onnx::Conv_580 = Identity(%onnx::Conv_571)
%onnx::Conv_577 = Identity(%onnx::Conv_574)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_570, %onnx::Conv_571)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_573, %onnx::Conv_574)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_576, %onnx::Conv_577)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_579, %onnx::Conv_580)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_582, %onnx::Conv_583)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_585, %onnx::Conv_586)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_588, %onnx::Conv_589)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_591, %onnx::Conv_592)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_594, %onnx::Conv_595)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_597, %onnx::Conv_598)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_600, %onnx::Conv_601)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_603, %onnx::Conv_604)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_606, %onnx::Conv_607)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.5/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.3/Add_output_0, %onnx::Conv_609, %onnx::Conv_610)
%/cells.5/relu/Relu_output_0 = Relu(%/cells.5/conv/Conv_output_0)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/relu/Relu_output_0, %onnx::Conv_612, %onnx::Conv_613)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_615, %onnx::Conv_616)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_618, %onnx::Conv_619)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/relu/Relu_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_621, %onnx::Conv_622)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_624, %onnx::Conv_625)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_627, %onnx::Conv_628)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_630, %onnx::Conv_631)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.15/Add_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%568 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %568
}
|
val_accuracy
| 0
| 45,291,008
| 1,385,692
|
{'zcp_synflow': 63.642632097272475, 'zcp_zen': 53.250404357910156, 'zcp_epe_nas': 10.932699472351937, 'zcp_fisher': 0.048367369920015335, 'zcp_flops': 45291008.0, 'zcp_grad_norm': 15.024925231933594, 'zcp_grasp': 0.006814002990722656, 'zcp_jacov': -16.06963136037693, 'zcp_l2_norm': 467.8146667480469, 'zcp_nwot': 208.3174544758455, 'zcp_params': 1385692.0, 'zcp_plain': -0.00017427206330467016, 'zcp_snip': 28.266847610473633, 'lat_1080ti_1': 0.24508549742516747, 'lat_1080ti_32': 0.29102526858773087, 'lat_1080ti_64': 0.2690532876484783, 'lat_2080ti_1': 0.3017903932338555, 'lat_2080ti_32': 0.2861004264511902, 'lat_2080ti_64': 0.2540555494502728, 'lat_essential_ph_1': 0.07547169811320754, 'lat_eyeriss': 0.1792859369556046, 'lat_fpga': 0.20697810190086968, 'lat_gold_6226': 0.1296143903009661, 'lat_gold_6240': 0.16712297266457327, 'lat_pixel2': 0.2608695652173913, 'lat_pixel3': 0.14196677903605356, 'lat_raspi4': 0.18636255368998644, 'lat_samsung_a50': 0.08421052631578947, 'lat_samsung_s7': 0.06299212598425197, 'lat_silver_4114': 0.15824075986566694, 'lat_silver_4210r': 0.1453422404771599, 'lat_titan_rtx_1': 0.2731326332962084, 'lat_titan_rtx_32': 0.2802289187047104, 'lat_titan_rtx_64': 0.24070418006846198, 'lat_titanx_1': 0.14490877921842685, 'lat_titanx_32': 0.25873746251871593, 'lat_titanx_64': 0.23628084221229817, 'lat_titanxp_1': 0.26595384835827196, 'lat_titanxp_32': 0.24792485198851325, 'lat_titanxp_64': 0.2409807254233323}
| |
FBNet_3228
|
FBNet
|
3228
|
3228
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_688[FLOAT, 16x3x3x3]
%onnx::Conv_689[FLOAT, 16]
%onnx::Conv_691[FLOAT, 96x16x1x1]
%onnx::Conv_692[FLOAT, 96]
%onnx::Conv_694[FLOAT, 96x1x3x3]
%onnx::Conv_697[FLOAT, 16x96x1x1]
%onnx::Conv_700[FLOAT, 96x16x1x1]
%onnx::Conv_703[FLOAT, 96x1x3x3]
%onnx::Conv_706[FLOAT, 24x96x1x1]
%onnx::Conv_707[FLOAT, 24]
%onnx::Conv_709[FLOAT, 72x24x1x1]
%onnx::Conv_710[FLOAT, 72]
%onnx::Conv_712[FLOAT, 72x1x5x5]
%onnx::Conv_715[FLOAT, 24x72x1x1]
%onnx::Conv_718[FLOAT, 24x12x1x1]
%onnx::Conv_721[FLOAT, 24x1x5x5]
%onnx::Conv_724[FLOAT, 24x12x1x1]
%onnx::Conv_727[FLOAT, 24x12x1x1]
%onnx::Conv_730[FLOAT, 24x1x3x3]
%onnx::Conv_733[FLOAT, 24x12x1x1]
%onnx::Conv_736[FLOAT, 144x24x1x1]
%onnx::Conv_737[FLOAT, 144]
%onnx::Conv_739[FLOAT, 144x1x3x3]
%onnx::Conv_742[FLOAT, 32x144x1x1]
%onnx::Conv_743[FLOAT, 32]
%onnx::Conv_745[FLOAT, 32x16x1x1]
%onnx::Conv_748[FLOAT, 32x1x3x3]
%onnx::Conv_751[FLOAT, 32x16x1x1]
%onnx::Conv_754[FLOAT, 192x32x1x1]
%onnx::Conv_755[FLOAT, 192]
%onnx::Conv_757[FLOAT, 192x1x5x5]
%onnx::Conv_760[FLOAT, 32x192x1x1]
%onnx::Conv_763[FLOAT, 192x32x1x1]
%onnx::Conv_766[FLOAT, 192x1x5x5]
%onnx::Conv_769[FLOAT, 32x192x1x1]
%onnx::Conv_772[FLOAT, 32x32x1x1]
%onnx::Conv_775[FLOAT, 32x1x5x5]
%onnx::Conv_778[FLOAT, 64x32x1x1]
%onnx::Conv_779[FLOAT, 64]
%onnx::Conv_781[FLOAT, 64x32x1x1]
%onnx::Conv_784[FLOAT, 64x1x5x5]
%onnx::Conv_787[FLOAT, 64x32x1x1]
%onnx::Conv_790[FLOAT, 192x64x1x1]
%onnx::Conv_793[FLOAT, 192x1x3x3]
%onnx::Conv_796[FLOAT, 64x192x1x1]
%onnx::Conv_799[FLOAT, 64x64x1x1]
%onnx::Conv_802[FLOAT, 64x1x5x5]
%onnx::Conv_805[FLOAT, 64x64x1x1]
%onnx::Conv_808[FLOAT, 112x64x1x1]
%onnx::Conv_809[FLOAT, 112]
%onnx::Conv_811[FLOAT, 336x112x1x1]
%onnx::Conv_812[FLOAT, 336]
%onnx::Conv_814[FLOAT, 336x1x3x3]
%onnx::Conv_817[FLOAT, 112x336x1x1]
%onnx::Conv_820[FLOAT, 112x56x1x1]
%onnx::Conv_823[FLOAT, 112x1x5x5]
%onnx::Conv_826[FLOAT, 112x56x1x1]
%onnx::Conv_829[FLOAT, 336x112x1x1]
%onnx::Conv_832[FLOAT, 336x1x3x3]
%onnx::Conv_835[FLOAT, 184x336x1x1]
%onnx::Conv_836[FLOAT, 184]
%onnx::Conv_838[FLOAT, 184x92x1x1]
%onnx::Conv_841[FLOAT, 184x1x5x5]
%onnx::Conv_844[FLOAT, 184x92x1x1]
%onnx::Conv_847[FLOAT, 1104x184x1x1]
%onnx::Conv_848[FLOAT, 1104]
%onnx::Conv_850[FLOAT, 1104x1x5x5]
%onnx::Conv_853[FLOAT, 184x1104x1x1]
%onnx::Conv_856[FLOAT, 1104x184x1x1]
%onnx::Conv_859[FLOAT, 1104x1x3x3]
%onnx::Conv_862[FLOAT, 184x1104x1x1]
%onnx::Conv_865[FLOAT, 184x184x1x1]
%onnx::Conv_868[FLOAT, 184x1x3x3]
%onnx::Conv_871[FLOAT, 352x184x1x1]
%onnx::Conv_872[FLOAT, 352]
%onnx::Conv_874[FLOAT, 1504x352x1x1]
%onnx::Conv_875[FLOAT, 1504]
) {
%onnx::Conv_869 = Identity(%onnx::Conv_836)
%onnx::Conv_866 = Identity(%onnx::Conv_836)
%onnx::Conv_863 = Identity(%onnx::Conv_836)
%onnx::Conv_860 = Identity(%onnx::Conv_848)
%onnx::Conv_857 = Identity(%onnx::Conv_848)
%onnx::Conv_854 = Identity(%onnx::Conv_836)
%onnx::Conv_851 = Identity(%onnx::Conv_848)
%onnx::Conv_845 = Identity(%onnx::Conv_836)
%onnx::Conv_842 = Identity(%onnx::Conv_836)
%onnx::Conv_839 = Identity(%onnx::Conv_836)
%onnx::Conv_833 = Identity(%onnx::Conv_812)
%onnx::Conv_830 = Identity(%onnx::Conv_812)
%onnx::Conv_827 = Identity(%onnx::Conv_809)
%onnx::Conv_824 = Identity(%onnx::Conv_809)
%onnx::Conv_821 = Identity(%onnx::Conv_809)
%onnx::Conv_818 = Identity(%onnx::Conv_809)
%onnx::Conv_815 = Identity(%onnx::Conv_812)
%onnx::Conv_806 = Identity(%onnx::Conv_779)
%onnx::Conv_803 = Identity(%onnx::Conv_779)
%onnx::Conv_800 = Identity(%onnx::Conv_779)
%onnx::Conv_797 = Identity(%onnx::Conv_779)
%onnx::Conv_794 = Identity(%onnx::Conv_755)
%onnx::Conv_791 = Identity(%onnx::Conv_755)
%onnx::Conv_788 = Identity(%onnx::Conv_779)
%onnx::Conv_785 = Identity(%onnx::Conv_779)
%onnx::Conv_782 = Identity(%onnx::Conv_779)
%onnx::Conv_776 = Identity(%onnx::Conv_743)
%onnx::Conv_773 = Identity(%onnx::Conv_743)
%onnx::Conv_770 = Identity(%onnx::Conv_743)
%onnx::Conv_767 = Identity(%onnx::Conv_755)
%onnx::Conv_764 = Identity(%onnx::Conv_755)
%onnx::Conv_761 = Identity(%onnx::Conv_743)
%onnx::Conv_758 = Identity(%onnx::Conv_755)
%onnx::Conv_752 = Identity(%onnx::Conv_743)
%onnx::Conv_749 = Identity(%onnx::Conv_743)
%onnx::Conv_746 = Identity(%onnx::Conv_743)
%onnx::Conv_740 = Identity(%onnx::Conv_737)
%onnx::Conv_734 = Identity(%onnx::Conv_707)
%onnx::Conv_731 = Identity(%onnx::Conv_707)
%onnx::Conv_728 = Identity(%onnx::Conv_707)
%onnx::Conv_725 = Identity(%onnx::Conv_707)
%onnx::Conv_722 = Identity(%onnx::Conv_707)
%onnx::Conv_719 = Identity(%onnx::Conv_707)
%onnx::Conv_716 = Identity(%onnx::Conv_707)
%onnx::Conv_713 = Identity(%onnx::Conv_710)
%onnx::Conv_704 = Identity(%onnx::Conv_692)
%onnx::Conv_701 = Identity(%onnx::Conv_692)
%onnx::Conv_698 = Identity(%onnx::Conv_689)
%onnx::Conv_695 = Identity(%onnx::Conv_692)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_688, %onnx::Conv_689)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.13/relu/Relu_output_0 = Relu(%/cells.13/conv/Conv_output_0)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/relu/Relu_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/relu/Relu_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_823, %onnx::Conv_824)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_826, %onnx::Conv_827)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_829, %onnx::Conv_830)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_832, %onnx::Conv_833)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_835, %onnx::Conv_836)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_838, %onnx::Conv_839)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_output_0 = Reshape(%/cells.18/nl/Relu_output_0, %/cells.18/shuffle/Constant_output_0)
%/cells.18/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.18/shuffle/Reshape_output_0)
%/cells.18/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_1_output_0 = Reshape(%/cells.18/shuffle/Transpose_output_0, %/cells.18/shuffle/Constant_1_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/shuffle/Reshape_1_output_0, %onnx::Conv_841, %onnx::Conv_842)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_844, %onnx::Conv_845)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_847, %onnx::Conv_848)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_850, %onnx::Conv_851)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_853, %onnx::Conv_854)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_856, %onnx::Conv_857)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_859, %onnx::Conv_860)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_862, %onnx::Conv_863)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_865, %onnx::Conv_866)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_868, %onnx::Conv_869)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_871, %onnx::Conv_872)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_874, %onnx::Conv_875)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%686 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %686
}
|
val_accuracy
| 0
| 71,907,712
| 2,004,028
|
{'zcp_synflow': 76.3164836909906, 'zcp_zen': 67.67366790771484, 'zcp_epe_nas': 11.205073802940477, 'zcp_fisher': 0.16555403172969818, 'zcp_flops': 71907712.0, 'zcp_grad_norm': 28.002517700195312, 'zcp_grasp': -0.2067108154296875, 'zcp_jacov': -16.062402875639087, 'zcp_l2_norm': 622.9871215820312, 'zcp_nwot': 215.63890297349172, 'zcp_params': 2004028.0, 'zcp_plain': -0.008964534848928452, 'zcp_snip': 50.84245681762695, 'lat_1080ti_1': 0.7072247232299234, 'lat_1080ti_32': 0.5822638864572173, 'lat_1080ti_64': 0.6125229979368468, 'lat_2080ti_1': 0.6974146813048149, 'lat_2080ti_32': 0.6350697999433188, 'lat_2080ti_64': 0.6207489704816193, 'lat_essential_ph_1': 0.22641509433962265, 'lat_eyeriss': 0.5471641024700639, 'lat_fpga': 0.4773251269060646, 'lat_gold_6226': 0.3950501371040084, 'lat_gold_6240': 0.5762373233287896, 'lat_pixel2': 0.43478260869565216, 'lat_pixel3': 0.5219837106159751, 'lat_raspi4': 0.5264904413981869, 'lat_samsung_a50': 0.23157894736842105, 'lat_samsung_s7': 0.2047244094488189, 'lat_silver_4114': 0.6032079291137875, 'lat_silver_4210r': 0.6231941268217907, 'lat_titan_rtx_1': 0.6446754418316282, 'lat_titan_rtx_32': 0.5838482135876875, 'lat_titan_rtx_64': 0.6306079760278317, 'lat_titanx_1': 0.3463606305646995, 'lat_titanx_32': 0.5966252683865578, 'lat_titanx_64': 0.6392126039440095, 'lat_titanxp_1': 0.6374904871072692, 'lat_titanxp_32': 0.6045116520566798, 'lat_titanxp_64': 0.6100447052674297}
| |
FBNet_1940
|
FBNet
|
1940
|
1940
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_706[FLOAT, 16x3x3x3]
%onnx::Conv_707[FLOAT, 16]
%onnx::Conv_709[FLOAT, 16x8x1x1]
%onnx::Conv_712[FLOAT, 16x1x5x5]
%onnx::Conv_715[FLOAT, 24x8x1x1]
%onnx::Conv_716[FLOAT, 24]
%onnx::Conv_718[FLOAT, 24x12x1x1]
%onnx::Conv_721[FLOAT, 24x1x5x5]
%onnx::Conv_724[FLOAT, 24x12x1x1]
%onnx::Conv_727[FLOAT, 24x12x1x1]
%onnx::Conv_730[FLOAT, 24x1x3x3]
%onnx::Conv_733[FLOAT, 24x12x1x1]
%onnx::Conv_736[FLOAT, 24x12x1x1]
%onnx::Conv_739[FLOAT, 24x1x3x3]
%onnx::Conv_742[FLOAT, 32x12x1x1]
%onnx::Conv_743[FLOAT, 32]
%onnx::Conv_745[FLOAT, 32x32x1x1]
%onnx::Conv_748[FLOAT, 32x1x3x3]
%onnx::Conv_751[FLOAT, 32x32x1x1]
%onnx::Conv_754[FLOAT, 96x32x1x1]
%onnx::Conv_755[FLOAT, 96]
%onnx::Conv_757[FLOAT, 96x1x5x5]
%onnx::Conv_760[FLOAT, 32x96x1x1]
%onnx::Conv_763[FLOAT, 32x16x1x1]
%onnx::Conv_766[FLOAT, 32x1x3x3]
%onnx::Conv_769[FLOAT, 32x16x1x1]
%onnx::Conv_772[FLOAT, 32x16x1x1]
%onnx::Conv_775[FLOAT, 32x1x5x5]
%onnx::Conv_778[FLOAT, 64x16x1x1]
%onnx::Conv_779[FLOAT, 64]
%onnx::Conv_781[FLOAT, 64x32x1x1]
%onnx::Conv_784[FLOAT, 64x1x3x3]
%onnx::Conv_787[FLOAT, 64x32x1x1]
%onnx::Conv_790[FLOAT, 192x64x1x1]
%onnx::Conv_791[FLOAT, 192]
%onnx::Conv_793[FLOAT, 192x1x3x3]
%onnx::Conv_796[FLOAT, 64x192x1x1]
%onnx::Conv_799[FLOAT, 192x64x1x1]
%onnx::Conv_802[FLOAT, 192x1x3x3]
%onnx::Conv_805[FLOAT, 64x192x1x1]
%onnx::Conv_808[FLOAT, 384x64x1x1]
%onnx::Conv_809[FLOAT, 384]
%onnx::Conv_811[FLOAT, 384x1x3x3]
%onnx::Conv_814[FLOAT, 112x384x1x1]
%onnx::Conv_815[FLOAT, 112]
%onnx::Conv_817[FLOAT, 336x112x1x1]
%onnx::Conv_818[FLOAT, 336]
%onnx::Conv_820[FLOAT, 336x1x3x3]
%onnx::Conv_823[FLOAT, 112x336x1x1]
%onnx::Conv_826[FLOAT, 112x56x1x1]
%onnx::Conv_829[FLOAT, 112x1x5x5]
%onnx::Conv_832[FLOAT, 112x56x1x1]
%onnx::Conv_835[FLOAT, 112x112x1x1]
%onnx::Conv_838[FLOAT, 112x1x3x3]
%onnx::Conv_841[FLOAT, 112x112x1x1]
%onnx::Conv_844[FLOAT, 112x112x1x1]
%onnx::Conv_847[FLOAT, 112x1x3x3]
%onnx::Conv_850[FLOAT, 184x112x1x1]
%onnx::Conv_851[FLOAT, 184]
%onnx::Conv_853[FLOAT, 184x92x1x1]
%onnx::Conv_856[FLOAT, 184x1x3x3]
%onnx::Conv_859[FLOAT, 184x92x1x1]
%onnx::Conv_862[FLOAT, 1104x184x1x1]
%onnx::Conv_863[FLOAT, 1104]
%onnx::Conv_865[FLOAT, 1104x1x5x5]
%onnx::Conv_868[FLOAT, 184x1104x1x1]
%onnx::Conv_871[FLOAT, 184x184x1x1]
%onnx::Conv_874[FLOAT, 184x1x5x5]
%onnx::Conv_877[FLOAT, 352x184x1x1]
%onnx::Conv_878[FLOAT, 352]
%onnx::Conv_880[FLOAT, 1504x352x1x1]
%onnx::Conv_881[FLOAT, 1504]
) {
%onnx::Conv_875 = Identity(%onnx::Conv_851)
%onnx::Conv_872 = Identity(%onnx::Conv_851)
%onnx::Conv_869 = Identity(%onnx::Conv_851)
%onnx::Conv_866 = Identity(%onnx::Conv_863)
%onnx::Conv_860 = Identity(%onnx::Conv_851)
%onnx::Conv_857 = Identity(%onnx::Conv_851)
%onnx::Conv_854 = Identity(%onnx::Conv_851)
%onnx::Conv_848 = Identity(%onnx::Conv_815)
%onnx::Conv_845 = Identity(%onnx::Conv_815)
%onnx::Conv_842 = Identity(%onnx::Conv_815)
%onnx::Conv_839 = Identity(%onnx::Conv_815)
%onnx::Conv_836 = Identity(%onnx::Conv_815)
%onnx::Conv_833 = Identity(%onnx::Conv_815)
%onnx::Conv_830 = Identity(%onnx::Conv_815)
%onnx::Conv_827 = Identity(%onnx::Conv_815)
%onnx::Conv_824 = Identity(%onnx::Conv_815)
%onnx::Conv_821 = Identity(%onnx::Conv_818)
%onnx::Conv_812 = Identity(%onnx::Conv_809)
%onnx::Conv_806 = Identity(%onnx::Conv_779)
%onnx::Conv_803 = Identity(%onnx::Conv_791)
%onnx::Conv_800 = Identity(%onnx::Conv_791)
%onnx::Conv_797 = Identity(%onnx::Conv_779)
%onnx::Conv_794 = Identity(%onnx::Conv_791)
%onnx::Conv_788 = Identity(%onnx::Conv_779)
%onnx::Conv_785 = Identity(%onnx::Conv_779)
%onnx::Conv_782 = Identity(%onnx::Conv_779)
%onnx::Conv_776 = Identity(%onnx::Conv_743)
%onnx::Conv_773 = Identity(%onnx::Conv_743)
%onnx::Conv_770 = Identity(%onnx::Conv_743)
%onnx::Conv_767 = Identity(%onnx::Conv_743)
%onnx::Conv_764 = Identity(%onnx::Conv_743)
%onnx::Conv_761 = Identity(%onnx::Conv_743)
%onnx::Conv_758 = Identity(%onnx::Conv_755)
%onnx::Conv_752 = Identity(%onnx::Conv_743)
%onnx::Conv_749 = Identity(%onnx::Conv_743)
%onnx::Conv_746 = Identity(%onnx::Conv_743)
%onnx::Conv_740 = Identity(%onnx::Conv_716)
%onnx::Conv_737 = Identity(%onnx::Conv_716)
%onnx::Conv_734 = Identity(%onnx::Conv_716)
%onnx::Conv_731 = Identity(%onnx::Conv_716)
%onnx::Conv_728 = Identity(%onnx::Conv_716)
%onnx::Conv_725 = Identity(%onnx::Conv_716)
%onnx::Conv_722 = Identity(%onnx::Conv_716)
%onnx::Conv_719 = Identity(%onnx::Conv_716)
%onnx::Conv_713 = Identity(%onnx::Conv_707)
%onnx::Conv_710 = Identity(%onnx::Conv_707)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_706, %onnx::Conv_707)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_823, %onnx::Conv_824)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_826, %onnx::Conv_827)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_829, %onnx::Conv_830)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_832, %onnx::Conv_833)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_835, %onnx::Conv_836)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_838, %onnx::Conv_839)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_841, %onnx::Conv_842)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_844, %onnx::Conv_845)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_847, %onnx::Conv_848)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_850, %onnx::Conv_851)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_853, %onnx::Conv_854)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_output_0 = Reshape(%/cells.18/nl/Relu_output_0, %/cells.18/shuffle/Constant_output_0)
%/cells.18/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.18/shuffle/Reshape_output_0)
%/cells.18/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_1_output_0 = Reshape(%/cells.18/shuffle/Transpose_output_0, %/cells.18/shuffle/Constant_1_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/shuffle/Reshape_1_output_0, %onnx::Conv_856, %onnx::Conv_857)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_859, %onnx::Conv_860)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_862, %onnx::Conv_863)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_865, %onnx::Conv_866)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_868, %onnx::Conv_869)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_871, %onnx::Conv_872)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_874, %onnx::Conv_875)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_877, %onnx::Conv_878)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_880, %onnx::Conv_881)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%704 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %704
}
|
val_accuracy
| 0
| 43,501,696
| 1,572,940
|
{'zcp_synflow': 65.99436738986599, 'zcp_zen': 59.80062484741211, 'zcp_epe_nas': 29.843358420072562, 'zcp_fisher': 0.045005545020103455, 'zcp_flops': 43501696.0, 'zcp_grad_norm': 16.040573120117188, 'zcp_grasp': -0.010781288146972656, 'zcp_jacov': -16.047763527648193, 'zcp_l2_norm': 526.12353515625, 'zcp_nwot': 201.20662908054186, 'zcp_params': 1572940.0, 'zcp_plain': 0.002944633597508073, 'zcp_snip': 28.45619010925293, 'lat_1080ti_1': 0.5512828535307734, 'lat_1080ti_32': 0.401061284975718, 'lat_1080ti_64': 0.19310733183628306, 'lat_2080ti_1': 0.5951814075134089, 'lat_2080ti_32': 0.45545819883458283, 'lat_2080ti_64': 0.2164065214748669, 'lat_essential_ph_1': 0.2641509433962264, 'lat_eyeriss': 0.14103386940017962, 'lat_fpga': 0.18163646978497708, 'lat_gold_6226': 0.20158953907952876, 'lat_gold_6240': 0.4456611300634854, 'lat_pixel2': 0.15217391304347827, 'lat_pixel3': 0.16782943972633182, 'lat_raspi4': 0.20174276158577562, 'lat_samsung_a50': 0.08421052631578947, 'lat_samsung_s7': 0.13385826771653545, 'lat_silver_4114': 0.39174181786605083, 'lat_silver_4210r': 0.4046145715273913, 'lat_titan_rtx_1': 0.5658394469838636, 'lat_titan_rtx_32': 0.4703305007636185, 'lat_titan_rtx_64': 0.3660859471867174, 'lat_titanx_1': 0.29236377586740075, 'lat_titanx_32': 0.33146345224377666, 'lat_titanx_64': 0.21303196020493342, 'lat_titanxp_1': 0.5545395801919245, 'lat_titanxp_32': 0.4004737359027265, 'lat_titanxp_64': 0.2199377818058232}
| |
FBNet_1416
|
FBNet
|
1416
|
1416
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_615[FLOAT, 16x3x3x3]
%onnx::Conv_616[FLOAT, 16]
%onnx::Conv_618[FLOAT, 16x16x1x1]
%onnx::Conv_621[FLOAT, 16x1x3x3]
%onnx::Conv_624[FLOAT, 16x16x1x1]
%onnx::Conv_627[FLOAT, 96x16x1x1]
%onnx::Conv_628[FLOAT, 96]
%onnx::Conv_630[FLOAT, 96x1x3x3]
%onnx::Conv_633[FLOAT, 24x96x1x1]
%onnx::Conv_634[FLOAT, 24]
%onnx::Conv_636[FLOAT, 24x12x1x1]
%onnx::Conv_639[FLOAT, 24x1x5x5]
%onnx::Conv_642[FLOAT, 24x12x1x1]
%onnx::Conv_645[FLOAT, 144x24x1x1]
%onnx::Conv_646[FLOAT, 144]
%onnx::Conv_648[FLOAT, 144x1x5x5]
%onnx::Conv_651[FLOAT, 24x144x1x1]
%onnx::Conv_654[FLOAT, 144x24x1x1]
%onnx::Conv_657[FLOAT, 144x1x3x3]
%onnx::Conv_660[FLOAT, 24x144x1x1]
%onnx::Conv_663[FLOAT, 24x24x1x1]
%onnx::Conv_666[FLOAT, 24x1x3x3]
%onnx::Conv_669[FLOAT, 32x24x1x1]
%onnx::Conv_670[FLOAT, 32]
%onnx::Conv_672[FLOAT, 96x32x1x1]
%onnx::Conv_675[FLOAT, 96x1x3x3]
%onnx::Conv_678[FLOAT, 32x96x1x1]
%onnx::Conv_681[FLOAT, 192x32x1x1]
%onnx::Conv_682[FLOAT, 192]
%onnx::Conv_684[FLOAT, 192x1x3x3]
%onnx::Conv_687[FLOAT, 32x192x1x1]
%onnx::Conv_690[FLOAT, 32x16x1x1]
%onnx::Conv_693[FLOAT, 32x1x5x5]
%onnx::Conv_696[FLOAT, 32x16x1x1]
%onnx::Conv_699[FLOAT, 64x32x1x1]
%onnx::Conv_700[FLOAT, 64]
%onnx::Conv_702[FLOAT, 64x64x1x1]
%onnx::Conv_705[FLOAT, 64x1x3x3]
%onnx::Conv_708[FLOAT, 64x64x1x1]
%onnx::Conv_711[FLOAT, 64x64x1x1]
%onnx::Conv_714[FLOAT, 64x1x5x5]
%onnx::Conv_717[FLOAT, 64x64x1x1]
%onnx::Conv_720[FLOAT, 192x64x1x1]
%onnx::Conv_723[FLOAT, 192x1x5x5]
%onnx::Conv_726[FLOAT, 64x192x1x1]
%onnx::Conv_729[FLOAT, 112x64x1x1]
%onnx::Conv_730[FLOAT, 112]
%onnx::Conv_732[FLOAT, 112x112x1x1]
%onnx::Conv_735[FLOAT, 112x1x5x5]
%onnx::Conv_738[FLOAT, 112x112x1x1]
%onnx::Conv_741[FLOAT, 112x56x1x1]
%onnx::Conv_744[FLOAT, 112x1x5x5]
%onnx::Conv_747[FLOAT, 112x56x1x1]
%onnx::Conv_750[FLOAT, 672x112x1x1]
%onnx::Conv_751[FLOAT, 672]
%onnx::Conv_753[FLOAT, 672x1x3x3]
%onnx::Conv_756[FLOAT, 112x672x1x1]
%onnx::Conv_759[FLOAT, 336x112x1x1]
%onnx::Conv_760[FLOAT, 336]
%onnx::Conv_762[FLOAT, 336x1x3x3]
%onnx::Conv_765[FLOAT, 184x336x1x1]
%onnx::Conv_766[FLOAT, 184]
%onnx::Conv_768[FLOAT, 184x184x1x1]
%onnx::Conv_771[FLOAT, 184x1x3x3]
%onnx::Conv_774[FLOAT, 184x184x1x1]
%onnx::Conv_777[FLOAT, 552x184x1x1]
%onnx::Conv_778[FLOAT, 552]
%onnx::Conv_780[FLOAT, 552x1x3x3]
%onnx::Conv_783[FLOAT, 184x552x1x1]
%onnx::Conv_786[FLOAT, 184x184x1x1]
%onnx::Conv_789[FLOAT, 184x1x5x5]
%onnx::Conv_792[FLOAT, 352x184x1x1]
%onnx::Conv_793[FLOAT, 352]
%onnx::Conv_795[FLOAT, 1504x352x1x1]
%onnx::Conv_796[FLOAT, 1504]
) {
%onnx::Conv_790 = Identity(%onnx::Conv_766)
%onnx::Conv_787 = Identity(%onnx::Conv_766)
%onnx::Conv_784 = Identity(%onnx::Conv_766)
%onnx::Conv_781 = Identity(%onnx::Conv_778)
%onnx::Conv_775 = Identity(%onnx::Conv_766)
%onnx::Conv_772 = Identity(%onnx::Conv_766)
%onnx::Conv_769 = Identity(%onnx::Conv_766)
%onnx::Conv_763 = Identity(%onnx::Conv_760)
%onnx::Conv_757 = Identity(%onnx::Conv_730)
%onnx::Conv_754 = Identity(%onnx::Conv_751)
%onnx::Conv_748 = Identity(%onnx::Conv_730)
%onnx::Conv_745 = Identity(%onnx::Conv_730)
%onnx::Conv_742 = Identity(%onnx::Conv_730)
%onnx::Conv_739 = Identity(%onnx::Conv_730)
%onnx::Conv_736 = Identity(%onnx::Conv_730)
%onnx::Conv_733 = Identity(%onnx::Conv_730)
%onnx::Conv_727 = Identity(%onnx::Conv_700)
%onnx::Conv_724 = Identity(%onnx::Conv_682)
%onnx::Conv_721 = Identity(%onnx::Conv_682)
%onnx::Conv_718 = Identity(%onnx::Conv_700)
%onnx::Conv_715 = Identity(%onnx::Conv_700)
%onnx::Conv_712 = Identity(%onnx::Conv_700)
%onnx::Conv_709 = Identity(%onnx::Conv_700)
%onnx::Conv_706 = Identity(%onnx::Conv_700)
%onnx::Conv_703 = Identity(%onnx::Conv_700)
%onnx::Conv_697 = Identity(%onnx::Conv_670)
%onnx::Conv_694 = Identity(%onnx::Conv_670)
%onnx::Conv_691 = Identity(%onnx::Conv_670)
%onnx::Conv_688 = Identity(%onnx::Conv_670)
%onnx::Conv_685 = Identity(%onnx::Conv_682)
%onnx::Conv_679 = Identity(%onnx::Conv_670)
%onnx::Conv_676 = Identity(%onnx::Conv_628)
%onnx::Conv_673 = Identity(%onnx::Conv_628)
%onnx::Conv_667 = Identity(%onnx::Conv_634)
%onnx::Conv_664 = Identity(%onnx::Conv_634)
%onnx::Conv_661 = Identity(%onnx::Conv_634)
%onnx::Conv_658 = Identity(%onnx::Conv_646)
%onnx::Conv_655 = Identity(%onnx::Conv_646)
%onnx::Conv_652 = Identity(%onnx::Conv_634)
%onnx::Conv_649 = Identity(%onnx::Conv_646)
%onnx::Conv_643 = Identity(%onnx::Conv_634)
%onnx::Conv_640 = Identity(%onnx::Conv_634)
%onnx::Conv_637 = Identity(%onnx::Conv_634)
%onnx::Conv_631 = Identity(%onnx::Conv_628)
%onnx::Conv_625 = Identity(%onnx::Conv_616)
%onnx::Conv_622 = Identity(%onnx::Conv_616)
%onnx::Conv_619 = Identity(%onnx::Conv_616)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_615, %onnx::Conv_616)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_618, %onnx::Conv_619)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_621, %onnx::Conv_622)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_624, %onnx::Conv_625)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_627, %onnx::Conv_628)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_630, %onnx::Conv_631)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.8/Add_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.9/relu/Relu_output_0 = Relu(%/cells.9/conv/Conv_output_0)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/relu/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/relu/Relu_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.13/relu/Relu_output_0 = Relu(%/cells.13/conv/Conv_output_0)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/relu/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/relu/Relu_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%613 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %613
}
|
val_accuracy
| 0
| 71,339,392
| 1,489,836
|
{'zcp_synflow': 78.15772579150546, 'zcp_zen': 65.2808837890625, 'zcp_epe_nas': 10.014863030655693, 'zcp_fisher': 0.13065938651561737, 'zcp_flops': 71339392.0, 'zcp_grad_norm': 25.021644592285156, 'zcp_grasp': -0.03489875793457031, 'zcp_jacov': -16.05583122650677, 'zcp_l2_norm': 588.1707763671875, 'zcp_nwot': 216.31108209209293, 'zcp_params': 1489836.0, 'zcp_plain': -0.003839142620563507, 'zcp_snip': 44.977535247802734, 'lat_1080ti_1': 0.5584926874027284, 'lat_1080ti_32': 0.61490214967519, 'lat_1080ti_64': 0.6112786802046803, 'lat_2080ti_1': 0.6433200856110352, 'lat_2080ti_32': 0.6476736033490369, 'lat_2080ti_64': 0.6169982991153371, 'lat_essential_ph_1': 0.22641509433962265, 'lat_eyeriss': 0.48433058881777996, 'lat_fpga': 0.4926033307275947, 'lat_gold_6226': 0.22724632834164493, 'lat_gold_6240': 0.4381533847004635, 'lat_pixel2': 0.41304347826086957, 'lat_pixel3': 0.4876302482771916, 'lat_raspi4': 0.47830630667218926, 'lat_samsung_a50': 0.17894736842105263, 'lat_samsung_s7': 0.14173228346456693, 'lat_silver_4114': 0.44296085269567104, 'lat_silver_4210r': 0.43858160147559405, 'lat_titan_rtx_1': 0.5315928359405884, 'lat_titan_rtx_32': 0.616111783341954, 'lat_titan_rtx_64': 0.6324061801128484, 'lat_titanx_1': 0.272109982390728, 'lat_titanx_32': 0.6103139666454728, 'lat_titanx_64': 0.6026665900976383, 'lat_titanxp_1': 0.49050074698811125, 'lat_titanxp_32': 0.6262216549099975, 'lat_titanxp_64': 0.6107979154665087}
| |
FBNet_950
|
FBNet
|
950
|
950
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_630[FLOAT, 16x3x3x3]
%onnx::Conv_631[FLOAT, 16]
%onnx::Conv_633[FLOAT, 16x8x1x1]
%onnx::Conv_636[FLOAT, 16x1x5x5]
%onnx::Conv_639[FLOAT, 16x8x1x1]
%onnx::Conv_642[FLOAT, 48x16x1x1]
%onnx::Conv_643[FLOAT, 48]
%onnx::Conv_645[FLOAT, 48x1x5x5]
%onnx::Conv_648[FLOAT, 24x48x1x1]
%onnx::Conv_649[FLOAT, 24]
%onnx::Conv_651[FLOAT, 144x24x1x1]
%onnx::Conv_652[FLOAT, 144]
%onnx::Conv_654[FLOAT, 144x1x5x5]
%onnx::Conv_657[FLOAT, 24x144x1x1]
%onnx::Conv_660[FLOAT, 72x24x1x1]
%onnx::Conv_661[FLOAT, 72]
%onnx::Conv_663[FLOAT, 72x1x3x3]
%onnx::Conv_666[FLOAT, 24x72x1x1]
%onnx::Conv_669[FLOAT, 24x24x1x1]
%onnx::Conv_672[FLOAT, 24x1x5x5]
%onnx::Conv_675[FLOAT, 24x24x1x1]
%onnx::Conv_678[FLOAT, 144x24x1x1]
%onnx::Conv_681[FLOAT, 144x1x3x3]
%onnx::Conv_684[FLOAT, 32x144x1x1]
%onnx::Conv_685[FLOAT, 32]
%onnx::Conv_687[FLOAT, 32x32x1x1]
%onnx::Conv_690[FLOAT, 32x1x5x5]
%onnx::Conv_693[FLOAT, 32x32x1x1]
%onnx::Conv_696[FLOAT, 32x32x1x1]
%onnx::Conv_699[FLOAT, 32x1x5x5]
%onnx::Conv_702[FLOAT, 32x32x1x1]
%onnx::Conv_705[FLOAT, 96x32x1x1]
%onnx::Conv_706[FLOAT, 96]
%onnx::Conv_708[FLOAT, 96x1x3x3]
%onnx::Conv_711[FLOAT, 32x96x1x1]
%onnx::Conv_714[FLOAT, 96x32x1x1]
%onnx::Conv_717[FLOAT, 96x1x5x5]
%onnx::Conv_720[FLOAT, 64x96x1x1]
%onnx::Conv_721[FLOAT, 64]
%onnx::Conv_723[FLOAT, 192x64x1x1]
%onnx::Conv_724[FLOAT, 192]
%onnx::Conv_726[FLOAT, 192x1x5x5]
%onnx::Conv_729[FLOAT, 64x192x1x1]
%onnx::Conv_732[FLOAT, 64x64x1x1]
%onnx::Conv_735[FLOAT, 64x1x3x3]
%onnx::Conv_738[FLOAT, 64x64x1x1]
%onnx::Conv_741[FLOAT, 192x64x1x1]
%onnx::Conv_744[FLOAT, 192x1x5x5]
%onnx::Conv_747[FLOAT, 64x192x1x1]
%onnx::Conv_750[FLOAT, 64x64x1x1]
%onnx::Conv_753[FLOAT, 64x1x3x3]
%onnx::Conv_756[FLOAT, 112x64x1x1]
%onnx::Conv_757[FLOAT, 112]
%onnx::Conv_759[FLOAT, 112x112x1x1]
%onnx::Conv_762[FLOAT, 112x1x3x3]
%onnx::Conv_765[FLOAT, 112x112x1x1]
%onnx::Conv_768[FLOAT, 672x112x1x1]
%onnx::Conv_769[FLOAT, 672]
%onnx::Conv_771[FLOAT, 672x1x3x3]
%onnx::Conv_774[FLOAT, 112x672x1x1]
%onnx::Conv_777[FLOAT, 112x112x1x1]
%onnx::Conv_780[FLOAT, 112x1x5x5]
%onnx::Conv_783[FLOAT, 112x112x1x1]
%onnx::Conv_786[FLOAT, 112x112x1x1]
%onnx::Conv_789[FLOAT, 112x1x3x3]
%onnx::Conv_792[FLOAT, 184x112x1x1]
%onnx::Conv_793[FLOAT, 184]
%onnx::Conv_795[FLOAT, 184x184x1x1]
%onnx::Conv_798[FLOAT, 184x1x5x5]
%onnx::Conv_801[FLOAT, 184x184x1x1]
%onnx::Conv_804[FLOAT, 552x184x1x1]
%onnx::Conv_805[FLOAT, 552]
%onnx::Conv_807[FLOAT, 552x1x3x3]
%onnx::Conv_810[FLOAT, 184x552x1x1]
%onnx::Conv_813[FLOAT, 184x92x1x1]
%onnx::Conv_816[FLOAT, 184x1x3x3]
%onnx::Conv_819[FLOAT, 352x92x1x1]
%onnx::Conv_820[FLOAT, 352]
%onnx::Conv_822[FLOAT, 1504x352x1x1]
%onnx::Conv_823[FLOAT, 1504]
) {
%onnx::Conv_817 = Identity(%onnx::Conv_793)
%onnx::Conv_814 = Identity(%onnx::Conv_793)
%onnx::Conv_811 = Identity(%onnx::Conv_793)
%onnx::Conv_808 = Identity(%onnx::Conv_805)
%onnx::Conv_802 = Identity(%onnx::Conv_793)
%onnx::Conv_799 = Identity(%onnx::Conv_793)
%onnx::Conv_796 = Identity(%onnx::Conv_793)
%onnx::Conv_790 = Identity(%onnx::Conv_757)
%onnx::Conv_787 = Identity(%onnx::Conv_757)
%onnx::Conv_784 = Identity(%onnx::Conv_757)
%onnx::Conv_781 = Identity(%onnx::Conv_757)
%onnx::Conv_778 = Identity(%onnx::Conv_757)
%onnx::Conv_775 = Identity(%onnx::Conv_757)
%onnx::Conv_772 = Identity(%onnx::Conv_769)
%onnx::Conv_766 = Identity(%onnx::Conv_757)
%onnx::Conv_763 = Identity(%onnx::Conv_757)
%onnx::Conv_760 = Identity(%onnx::Conv_757)
%onnx::Conv_754 = Identity(%onnx::Conv_721)
%onnx::Conv_751 = Identity(%onnx::Conv_721)
%onnx::Conv_748 = Identity(%onnx::Conv_721)
%onnx::Conv_745 = Identity(%onnx::Conv_724)
%onnx::Conv_742 = Identity(%onnx::Conv_724)
%onnx::Conv_739 = Identity(%onnx::Conv_721)
%onnx::Conv_736 = Identity(%onnx::Conv_721)
%onnx::Conv_733 = Identity(%onnx::Conv_721)
%onnx::Conv_730 = Identity(%onnx::Conv_721)
%onnx::Conv_727 = Identity(%onnx::Conv_724)
%onnx::Conv_718 = Identity(%onnx::Conv_706)
%onnx::Conv_715 = Identity(%onnx::Conv_706)
%onnx::Conv_712 = Identity(%onnx::Conv_685)
%onnx::Conv_709 = Identity(%onnx::Conv_706)
%onnx::Conv_703 = Identity(%onnx::Conv_685)
%onnx::Conv_700 = Identity(%onnx::Conv_685)
%onnx::Conv_697 = Identity(%onnx::Conv_685)
%onnx::Conv_694 = Identity(%onnx::Conv_685)
%onnx::Conv_691 = Identity(%onnx::Conv_685)
%onnx::Conv_688 = Identity(%onnx::Conv_685)
%onnx::Conv_682 = Identity(%onnx::Conv_652)
%onnx::Conv_679 = Identity(%onnx::Conv_652)
%onnx::Conv_676 = Identity(%onnx::Conv_649)
%onnx::Conv_673 = Identity(%onnx::Conv_649)
%onnx::Conv_670 = Identity(%onnx::Conv_649)
%onnx::Conv_667 = Identity(%onnx::Conv_649)
%onnx::Conv_664 = Identity(%onnx::Conv_661)
%onnx::Conv_658 = Identity(%onnx::Conv_649)
%onnx::Conv_655 = Identity(%onnx::Conv_652)
%onnx::Conv_646 = Identity(%onnx::Conv_643)
%onnx::Conv_640 = Identity(%onnx::Conv_631)
%onnx::Conv_637 = Identity(%onnx::Conv_631)
%onnx::Conv_634 = Identity(%onnx::Conv_631)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_630, %onnx::Conv_631)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%628 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %628
}
|
val_accuracy
| 0
| 67,669,888
| 1,408,940
|
{'zcp_synflow': 83.27929248421324, 'zcp_zen': 69.97463989257812, 'zcp_epe_nas': 29.697726265159904, 'zcp_fisher': 0.1593855619430542, 'zcp_flops': 67669888.0, 'zcp_grad_norm': 25.683147430419922, 'zcp_grasp': -0.209991455078125, 'zcp_jacov': -16.152415797322302, 'zcp_l2_norm': 606.703125, 'zcp_nwot': 214.29595474720807, 'zcp_params': 1408940.0, 'zcp_plain': -0.00036837393417954445, 'zcp_snip': 44.476139068603516, 'lat_1080ti_1': 0.6090494812801542, 'lat_1080ti_32': 0.6023298189910614, 'lat_1080ti_64': 0.5471018844555409, 'lat_2080ti_1': 0.7087371822807429, 'lat_2080ti_32': 0.5994804617608919, 'lat_2080ti_64': 0.5811684747329202, 'lat_essential_ph_1': 0.3584905660377358, 'lat_eyeriss': 0.42687913740304034, 'lat_fpga': 0.4143718026103563, 'lat_gold_6226': 0.21374326078836783, 'lat_gold_6240': 0.358445337148733, 'lat_pixel2': 0.32608695652173914, 'lat_pixel3': 0.44260804050577945, 'lat_raspi4': 0.4419973094392004, 'lat_samsung_a50': 0.15789473684210525, 'lat_samsung_s7': 0.18110236220472442, 'lat_silver_4114': 0.36747324742724125, 'lat_silver_4210r': 0.3841412304876875, 'lat_titan_rtx_1': 0.6129259683664781, 'lat_titan_rtx_32': 0.5822133705457959, 'lat_titan_rtx_64': 0.6003674610789942, 'lat_titanx_1': 0.33131728120587606, 'lat_titanx_32': 0.6050128286268364, 'lat_titanx_64': 0.5479261554121584, 'lat_titanxp_1': 0.5757866162658261, 'lat_titanxp_32': 0.6310978048271657, 'lat_titanxp_64': 0.5961475541320818}
| |
FBNet_712
|
FBNet
|
712
|
712
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_586[FLOAT, 16x3x3x3]
%onnx::Conv_587[FLOAT, 16]
%onnx::Conv_589[FLOAT, 96x16x1x1]
%onnx::Conv_590[FLOAT, 96]
%onnx::Conv_592[FLOAT, 96x1x5x5]
%onnx::Conv_595[FLOAT, 16x96x1x1]
%onnx::Conv_598[FLOAT, 16x8x1x1]
%onnx::Conv_601[FLOAT, 16x1x3x3]
%onnx::Conv_604[FLOAT, 24x8x1x1]
%onnx::Conv_605[FLOAT, 24]
%onnx::Conv_607[FLOAT, 24x24x1x1]
%onnx::Conv_610[FLOAT, 24x1x3x3]
%onnx::Conv_613[FLOAT, 24x24x1x1]
%onnx::Conv_616[FLOAT, 72x24x1x1]
%onnx::Conv_617[FLOAT, 72]
%onnx::Conv_619[FLOAT, 72x1x3x3]
%onnx::Conv_622[FLOAT, 24x72x1x1]
%onnx::Conv_625[FLOAT, 72x24x1x1]
%onnx::Conv_628[FLOAT, 72x1x3x3]
%onnx::Conv_631[FLOAT, 32x72x1x1]
%onnx::Conv_632[FLOAT, 32]
%onnx::Conv_634[FLOAT, 192x32x1x1]
%onnx::Conv_635[FLOAT, 192]
%onnx::Conv_637[FLOAT, 192x1x3x3]
%onnx::Conv_640[FLOAT, 32x192x1x1]
%onnx::Conv_643[FLOAT, 32x32x1x1]
%onnx::Conv_646[FLOAT, 32x1x3x3]
%onnx::Conv_649[FLOAT, 32x32x1x1]
%onnx::Conv_652[FLOAT, 192x32x1x1]
%onnx::Conv_655[FLOAT, 192x1x5x5]
%onnx::Conv_658[FLOAT, 32x192x1x1]
%onnx::Conv_661[FLOAT, 32x32x1x1]
%onnx::Conv_664[FLOAT, 32x1x3x3]
%onnx::Conv_667[FLOAT, 64x32x1x1]
%onnx::Conv_668[FLOAT, 64]
%onnx::Conv_670[FLOAT, 192x64x1x1]
%onnx::Conv_673[FLOAT, 192x1x5x5]
%onnx::Conv_676[FLOAT, 64x192x1x1]
%onnx::Conv_679[FLOAT, 64x32x1x1]
%onnx::Conv_682[FLOAT, 64x1x3x3]
%onnx::Conv_685[FLOAT, 64x32x1x1]
%onnx::Conv_688[FLOAT, 64x64x1x1]
%onnx::Conv_691[FLOAT, 64x1x5x5]
%onnx::Conv_694[FLOAT, 64x64x1x1]
%onnx::Conv_697[FLOAT, 384x64x1x1]
%onnx::Conv_698[FLOAT, 384]
%onnx::Conv_700[FLOAT, 384x1x5x5]
%onnx::Conv_703[FLOAT, 112x384x1x1]
%onnx::Conv_704[FLOAT, 112]
%onnx::Conv_706[FLOAT, 336x112x1x1]
%onnx::Conv_707[FLOAT, 336]
%onnx::Conv_709[FLOAT, 336x1x5x5]
%onnx::Conv_712[FLOAT, 112x336x1x1]
%onnx::Conv_715[FLOAT, 112x112x1x1]
%onnx::Conv_718[FLOAT, 112x1x5x5]
%onnx::Conv_721[FLOAT, 184x112x1x1]
%onnx::Conv_722[FLOAT, 184]
%onnx::Conv_724[FLOAT, 184x92x1x1]
%onnx::Conv_727[FLOAT, 184x1x5x5]
%onnx::Conv_730[FLOAT, 184x92x1x1]
%onnx::Conv_733[FLOAT, 184x92x1x1]
%onnx::Conv_736[FLOAT, 184x1x5x5]
%onnx::Conv_739[FLOAT, 184x92x1x1]
%onnx::Conv_742[FLOAT, 552x184x1x1]
%onnx::Conv_743[FLOAT, 552]
%onnx::Conv_745[FLOAT, 552x1x3x3]
%onnx::Conv_748[FLOAT, 352x552x1x1]
%onnx::Conv_749[FLOAT, 352]
%onnx::Conv_751[FLOAT, 1504x352x1x1]
%onnx::Conv_752[FLOAT, 1504]
) {
%onnx::Conv_746 = Identity(%onnx::Conv_743)
%onnx::Conv_740 = Identity(%onnx::Conv_722)
%onnx::Conv_737 = Identity(%onnx::Conv_722)
%onnx::Conv_734 = Identity(%onnx::Conv_722)
%onnx::Conv_731 = Identity(%onnx::Conv_722)
%onnx::Conv_728 = Identity(%onnx::Conv_722)
%onnx::Conv_725 = Identity(%onnx::Conv_722)
%onnx::Conv_719 = Identity(%onnx::Conv_704)
%onnx::Conv_716 = Identity(%onnx::Conv_704)
%onnx::Conv_713 = Identity(%onnx::Conv_704)
%onnx::Conv_710 = Identity(%onnx::Conv_707)
%onnx::Conv_701 = Identity(%onnx::Conv_698)
%onnx::Conv_695 = Identity(%onnx::Conv_668)
%onnx::Conv_692 = Identity(%onnx::Conv_668)
%onnx::Conv_689 = Identity(%onnx::Conv_668)
%onnx::Conv_686 = Identity(%onnx::Conv_668)
%onnx::Conv_683 = Identity(%onnx::Conv_668)
%onnx::Conv_680 = Identity(%onnx::Conv_668)
%onnx::Conv_677 = Identity(%onnx::Conv_668)
%onnx::Conv_674 = Identity(%onnx::Conv_635)
%onnx::Conv_671 = Identity(%onnx::Conv_635)
%onnx::Conv_665 = Identity(%onnx::Conv_632)
%onnx::Conv_662 = Identity(%onnx::Conv_632)
%onnx::Conv_659 = Identity(%onnx::Conv_632)
%onnx::Conv_656 = Identity(%onnx::Conv_635)
%onnx::Conv_653 = Identity(%onnx::Conv_635)
%onnx::Conv_650 = Identity(%onnx::Conv_632)
%onnx::Conv_647 = Identity(%onnx::Conv_632)
%onnx::Conv_644 = Identity(%onnx::Conv_632)
%onnx::Conv_641 = Identity(%onnx::Conv_632)
%onnx::Conv_638 = Identity(%onnx::Conv_635)
%onnx::Conv_629 = Identity(%onnx::Conv_617)
%onnx::Conv_626 = Identity(%onnx::Conv_617)
%onnx::Conv_623 = Identity(%onnx::Conv_605)
%onnx::Conv_620 = Identity(%onnx::Conv_617)
%onnx::Conv_614 = Identity(%onnx::Conv_605)
%onnx::Conv_611 = Identity(%onnx::Conv_605)
%onnx::Conv_608 = Identity(%onnx::Conv_605)
%onnx::Conv_602 = Identity(%onnx::Conv_587)
%onnx::Conv_599 = Identity(%onnx::Conv_587)
%onnx::Conv_596 = Identity(%onnx::Conv_587)
%onnx::Conv_593 = Identity(%onnx::Conv_590)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_586, %onnx::Conv_587)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_589, %onnx::Conv_590)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_592, %onnx::Conv_593)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_595, %onnx::Conv_596)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_598, %onnx::Conv_599)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_601, %onnx::Conv_602)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_604, %onnx::Conv_605)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_607, %onnx::Conv_608)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_610, %onnx::Conv_611)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_613, %onnx::Conv_614)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_616, %onnx::Conv_617)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_619, %onnx::Conv_620)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_622, %onnx::Conv_623)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_625, %onnx::Conv_626)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_628, %onnx::Conv_629)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_631, %onnx::Conv_632)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_634, %onnx::Conv_635)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_637, %onnx::Conv_638)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_640, %onnx::Conv_641)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_643, %onnx::Conv_644)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_646, %onnx::Conv_647)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_649, %onnx::Conv_650)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_652, %onnx::Conv_653)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_655, %onnx::Conv_656)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_658, %onnx::Conv_659)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%584 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %584
}
|
val_accuracy
| 0
| 55,294,336
| 1,369,060
|
{'zcp_synflow': 68.48874841828638, 'zcp_zen': 58.737953186035156, 'zcp_epe_nas': 21.69454797922005, 'zcp_fisher': 0.07012871652841568, 'zcp_flops': 55294336.0, 'zcp_grad_norm': 19.53650665283203, 'zcp_grasp': -0.04094886779785156, 'zcp_jacov': -16.04562632611698, 'zcp_l2_norm': 506.8324279785156, 'zcp_nwot': 210.76563260492827, 'zcp_params': 1369060.0, 'zcp_plain': -0.006965348031371832, 'zcp_snip': 32.76047897338867, 'lat_1080ti_1': 0.3164325883859177, 'lat_1080ti_32': 0.34843311914885605, 'lat_1080ti_64': 0.3352483890482033, 'lat_2080ti_1': 0.37627350583426755, 'lat_2080ti_32': 0.31668097820898716, 'lat_2080ti_64': 0.3192767303263862, 'lat_essential_ph_1': 0.2830188679245283, 'lat_eyeriss': 0.29780133130398123, 'lat_fpga': 0.29012754925340634, 'lat_gold_6226': 0.18559211344926685, 'lat_gold_6240': 0.3011582678176363, 'lat_pixel2': 0.15217391304347827, 'lat_pixel3': 0.30758509988984145, 'lat_raspi4': 0.33225723092579074, 'lat_samsung_a50': 0.10526315789473684, 'lat_samsung_s7': 0.12598425196850394, 'lat_silver_4114': 0.2611503704739714, 'lat_silver_4210r': 0.1971475490211282, 'lat_titan_rtx_1': 0.3681068441386087, 'lat_titan_rtx_32': 0.3234608141449821, 'lat_titan_rtx_64': 0.3114281785797521, 'lat_titanx_1': 0.191741366833574, 'lat_titanx_32': 0.3054676207844121, 'lat_titanx_64': 0.31686046290649994, 'lat_titanxp_1': 0.3354391473262753, 'lat_titanxp_32': 0.3327723886318919, 'lat_titanxp_64': 0.30496870746967736}
| |
FBNet_2387
|
FBNet
|
2387
|
2387
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_686[FLOAT, 16x3x3x3]
%onnx::Conv_687[FLOAT, 16]
%onnx::Conv_689[FLOAT, 16x8x1x1]
%onnx::Conv_692[FLOAT, 16x1x5x5]
%onnx::Conv_695[FLOAT, 16x8x1x1]
%onnx::Conv_698[FLOAT, 16x8x1x1]
%onnx::Conv_701[FLOAT, 16x1x5x5]
%onnx::Conv_704[FLOAT, 24x8x1x1]
%onnx::Conv_705[FLOAT, 24]
%onnx::Conv_707[FLOAT, 24x24x1x1]
%onnx::Conv_710[FLOAT, 24x1x5x5]
%onnx::Conv_713[FLOAT, 24x24x1x1]
%onnx::Conv_716[FLOAT, 24x24x1x1]
%onnx::Conv_719[FLOAT, 24x1x5x5]
%onnx::Conv_722[FLOAT, 24x24x1x1]
%onnx::Conv_725[FLOAT, 72x24x1x1]
%onnx::Conv_726[FLOAT, 72]
%onnx::Conv_728[FLOAT, 72x1x3x3]
%onnx::Conv_731[FLOAT, 24x72x1x1]
%onnx::Conv_734[FLOAT, 72x24x1x1]
%onnx::Conv_737[FLOAT, 72x1x3x3]
%onnx::Conv_740[FLOAT, 32x72x1x1]
%onnx::Conv_741[FLOAT, 32]
%onnx::Conv_743[FLOAT, 32x16x1x1]
%onnx::Conv_746[FLOAT, 32x1x5x5]
%onnx::Conv_749[FLOAT, 32x16x1x1]
%onnx::Conv_752[FLOAT, 96x32x1x1]
%onnx::Conv_753[FLOAT, 96]
%onnx::Conv_755[FLOAT, 96x1x3x3]
%onnx::Conv_758[FLOAT, 32x96x1x1]
%onnx::Conv_761[FLOAT, 192x32x1x1]
%onnx::Conv_762[FLOAT, 192]
%onnx::Conv_764[FLOAT, 192x1x3x3]
%onnx::Conv_767[FLOAT, 32x192x1x1]
%onnx::Conv_770[FLOAT, 192x32x1x1]
%onnx::Conv_773[FLOAT, 192x1x5x5]
%onnx::Conv_776[FLOAT, 64x192x1x1]
%onnx::Conv_777[FLOAT, 64]
%onnx::Conv_779[FLOAT, 64x64x1x1]
%onnx::Conv_782[FLOAT, 64x1x3x3]
%onnx::Conv_785[FLOAT, 64x64x1x1]
%onnx::Conv_788[FLOAT, 192x64x1x1]
%onnx::Conv_791[FLOAT, 192x1x5x5]
%onnx::Conv_794[FLOAT, 64x192x1x1]
%onnx::Conv_797[FLOAT, 384x64x1x1]
%onnx::Conv_798[FLOAT, 384]
%onnx::Conv_800[FLOAT, 384x1x5x5]
%onnx::Conv_803[FLOAT, 64x384x1x1]
%onnx::Conv_806[FLOAT, 192x64x1x1]
%onnx::Conv_809[FLOAT, 192x1x3x3]
%onnx::Conv_812[FLOAT, 112x192x1x1]
%onnx::Conv_813[FLOAT, 112]
%onnx::Conv_815[FLOAT, 672x112x1x1]
%onnx::Conv_816[FLOAT, 672]
%onnx::Conv_818[FLOAT, 672x1x3x3]
%onnx::Conv_821[FLOAT, 112x672x1x1]
%onnx::Conv_824[FLOAT, 672x112x1x1]
%onnx::Conv_827[FLOAT, 672x1x3x3]
%onnx::Conv_830[FLOAT, 112x672x1x1]
%onnx::Conv_833[FLOAT, 112x56x1x1]
%onnx::Conv_836[FLOAT, 112x1x3x3]
%onnx::Conv_839[FLOAT, 184x56x1x1]
%onnx::Conv_840[FLOAT, 184]
%onnx::Conv_842[FLOAT, 1104x184x1x1]
%onnx::Conv_843[FLOAT, 1104]
%onnx::Conv_845[FLOAT, 1104x1x5x5]
%onnx::Conv_848[FLOAT, 184x1104x1x1]
%onnx::Conv_851[FLOAT, 184x92x1x1]
%onnx::Conv_854[FLOAT, 184x1x3x3]
%onnx::Conv_857[FLOAT, 184x92x1x1]
%onnx::Conv_860[FLOAT, 1104x184x1x1]
%onnx::Conv_863[FLOAT, 1104x1x5x5]
%onnx::Conv_866[FLOAT, 184x1104x1x1]
%onnx::Conv_869[FLOAT, 1104x184x1x1]
%onnx::Conv_872[FLOAT, 1104x1x5x5]
%onnx::Conv_875[FLOAT, 352x1104x1x1]
%onnx::Conv_876[FLOAT, 352]
%onnx::Conv_878[FLOAT, 1504x352x1x1]
%onnx::Conv_879[FLOAT, 1504]
) {
%onnx::Conv_873 = Identity(%onnx::Conv_843)
%onnx::Conv_870 = Identity(%onnx::Conv_843)
%onnx::Conv_867 = Identity(%onnx::Conv_840)
%onnx::Conv_864 = Identity(%onnx::Conv_843)
%onnx::Conv_861 = Identity(%onnx::Conv_843)
%onnx::Conv_858 = Identity(%onnx::Conv_840)
%onnx::Conv_855 = Identity(%onnx::Conv_840)
%onnx::Conv_852 = Identity(%onnx::Conv_840)
%onnx::Conv_849 = Identity(%onnx::Conv_840)
%onnx::Conv_846 = Identity(%onnx::Conv_843)
%onnx::Conv_837 = Identity(%onnx::Conv_813)
%onnx::Conv_834 = Identity(%onnx::Conv_813)
%onnx::Conv_831 = Identity(%onnx::Conv_813)
%onnx::Conv_828 = Identity(%onnx::Conv_816)
%onnx::Conv_825 = Identity(%onnx::Conv_816)
%onnx::Conv_822 = Identity(%onnx::Conv_813)
%onnx::Conv_819 = Identity(%onnx::Conv_816)
%onnx::Conv_810 = Identity(%onnx::Conv_762)
%onnx::Conv_807 = Identity(%onnx::Conv_762)
%onnx::Conv_804 = Identity(%onnx::Conv_777)
%onnx::Conv_801 = Identity(%onnx::Conv_798)
%onnx::Conv_795 = Identity(%onnx::Conv_777)
%onnx::Conv_792 = Identity(%onnx::Conv_762)
%onnx::Conv_789 = Identity(%onnx::Conv_762)
%onnx::Conv_786 = Identity(%onnx::Conv_777)
%onnx::Conv_783 = Identity(%onnx::Conv_777)
%onnx::Conv_780 = Identity(%onnx::Conv_777)
%onnx::Conv_774 = Identity(%onnx::Conv_762)
%onnx::Conv_771 = Identity(%onnx::Conv_762)
%onnx::Conv_768 = Identity(%onnx::Conv_741)
%onnx::Conv_765 = Identity(%onnx::Conv_762)
%onnx::Conv_759 = Identity(%onnx::Conv_741)
%onnx::Conv_756 = Identity(%onnx::Conv_753)
%onnx::Conv_750 = Identity(%onnx::Conv_741)
%onnx::Conv_747 = Identity(%onnx::Conv_741)
%onnx::Conv_744 = Identity(%onnx::Conv_741)
%onnx::Conv_738 = Identity(%onnx::Conv_726)
%onnx::Conv_735 = Identity(%onnx::Conv_726)
%onnx::Conv_732 = Identity(%onnx::Conv_705)
%onnx::Conv_729 = Identity(%onnx::Conv_726)
%onnx::Conv_723 = Identity(%onnx::Conv_705)
%onnx::Conv_720 = Identity(%onnx::Conv_705)
%onnx::Conv_717 = Identity(%onnx::Conv_705)
%onnx::Conv_714 = Identity(%onnx::Conv_705)
%onnx::Conv_711 = Identity(%onnx::Conv_705)
%onnx::Conv_708 = Identity(%onnx::Conv_705)
%onnx::Conv_702 = Identity(%onnx::Conv_687)
%onnx::Conv_699 = Identity(%onnx::Conv_687)
%onnx::Conv_696 = Identity(%onnx::Conv_687)
%onnx::Conv_693 = Identity(%onnx::Conv_687)
%onnx::Conv_690 = Identity(%onnx::Conv_687)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_686, %onnx::Conv_687)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_803, %onnx::Conv_804)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_806, %onnx::Conv_807)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_809, %onnx::Conv_810)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_812, %onnx::Conv_813)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_815, %onnx::Conv_816)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_818, %onnx::Conv_819)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_821, %onnx::Conv_822)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_824, %onnx::Conv_825)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_827, %onnx::Conv_828)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_830, %onnx::Conv_831)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_833, %onnx::Conv_834)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_836, %onnx::Conv_837)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_839, %onnx::Conv_840)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_842, %onnx::Conv_843)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_845, %onnx::Conv_846)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_848, %onnx::Conv_849)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_851, %onnx::Conv_852)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_854, %onnx::Conv_855)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_857, %onnx::Conv_858)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_860, %onnx::Conv_861)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_863, %onnx::Conv_864)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_866, %onnx::Conv_867)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_869, %onnx::Conv_870)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_872, %onnx::Conv_873)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_875, %onnx::Conv_876)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_878, %onnx::Conv_879)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%684 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %684
}
|
val_accuracy
| 0
| 85,836,416
| 2,758,796
|
{'zcp_synflow': 79.05642482308495, 'zcp_zen': 72.05130004882812, 'zcp_epe_nas': 15.903563669543958, 'zcp_fisher': 0.11531393975019455, 'zcp_flops': 85836416.0, 'zcp_grad_norm': 26.697450637817383, 'zcp_grasp': -0.09046363830566406, 'zcp_jacov': -16.054494270186282, 'zcp_l2_norm': 702.1881713867188, 'zcp_nwot': 212.57259328055045, 'zcp_params': 2758796.0, 'zcp_plain': -0.0016807266511023045, 'zcp_snip': 44.818809509277344, 'lat_1080ti_1': 0.6781783834658457, 'lat_1080ti_32': 0.4951741958233866, 'lat_1080ti_64': 0.411194772060115, 'lat_2080ti_1': 0.7153794915721869, 'lat_2080ti_32': 0.5460187488773811, 'lat_2080ti_64': 0.4214634628213312, 'lat_essential_ph_1': 0.6226415094339622, 'lat_eyeriss': 0.6162507105789156, 'lat_fpga': 0.7423089482381579, 'lat_gold_6226': 0.627726464651286, 'lat_gold_6240': 0.8268132996015735, 'lat_pixel2': 0.45652173913043476, 'lat_pixel3': 0.5871010192744622, 'lat_raspi4': 0.6970256646731349, 'lat_samsung_a50': 0.30526315789473685, 'lat_samsung_s7': 0.2677165354330709, 'lat_silver_4114': 0.863271884917774, 'lat_silver_4210r': 0.856972875164298, 'lat_titan_rtx_1': 0.6847746283384508, 'lat_titan_rtx_32': 0.5622142708839133, 'lat_titan_rtx_64': 0.4163192183277046, 'lat_titanx_1': 0.3723237537837205, 'lat_titanx_32': 0.4720304556693366, 'lat_titanx_64': 0.3966098473893691, 'lat_titanxp_1': 0.6576646397829223, 'lat_titanxp_32': 0.5162566285863468, 'lat_titanxp_64': 0.4149594214247507}
| |
FBNet_4538
|
FBNet
|
4538
|
4538
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_723[FLOAT, 16x3x3x3]
%onnx::Conv_724[FLOAT, 16]
%onnx::Conv_726[FLOAT, 96x16x1x1]
%onnx::Conv_727[FLOAT, 96]
%onnx::Conv_729[FLOAT, 96x1x5x5]
%onnx::Conv_732[FLOAT, 16x96x1x1]
%onnx::Conv_735[FLOAT, 16x8x1x1]
%onnx::Conv_738[FLOAT, 16x1x3x3]
%onnx::Conv_741[FLOAT, 24x8x1x1]
%onnx::Conv_742[FLOAT, 24]
%onnx::Conv_744[FLOAT, 144x24x1x1]
%onnx::Conv_745[FLOAT, 144]
%onnx::Conv_747[FLOAT, 144x1x5x5]
%onnx::Conv_750[FLOAT, 24x144x1x1]
%onnx::Conv_753[FLOAT, 144x24x1x1]
%onnx::Conv_756[FLOAT, 144x1x3x3]
%onnx::Conv_759[FLOAT, 24x144x1x1]
%onnx::Conv_762[FLOAT, 24x24x1x1]
%onnx::Conv_765[FLOAT, 24x1x5x5]
%onnx::Conv_768[FLOAT, 24x24x1x1]
%onnx::Conv_771[FLOAT, 144x24x1x1]
%onnx::Conv_774[FLOAT, 144x1x5x5]
%onnx::Conv_777[FLOAT, 32x144x1x1]
%onnx::Conv_778[FLOAT, 32]
%onnx::Conv_780[FLOAT, 32x16x1x1]
%onnx::Conv_783[FLOAT, 32x1x5x5]
%onnx::Conv_786[FLOAT, 32x16x1x1]
%onnx::Conv_789[FLOAT, 192x32x1x1]
%onnx::Conv_790[FLOAT, 192]
%onnx::Conv_792[FLOAT, 192x1x5x5]
%onnx::Conv_795[FLOAT, 32x192x1x1]
%onnx::Conv_798[FLOAT, 192x32x1x1]
%onnx::Conv_801[FLOAT, 192x1x3x3]
%onnx::Conv_804[FLOAT, 32x192x1x1]
%onnx::Conv_807[FLOAT, 192x32x1x1]
%onnx::Conv_810[FLOAT, 192x1x3x3]
%onnx::Conv_813[FLOAT, 64x192x1x1]
%onnx::Conv_814[FLOAT, 64]
%onnx::Conv_816[FLOAT, 64x32x1x1]
%onnx::Conv_819[FLOAT, 64x1x3x3]
%onnx::Conv_822[FLOAT, 64x32x1x1]
%onnx::Conv_825[FLOAT, 192x64x1x1]
%onnx::Conv_828[FLOAT, 192x1x3x3]
%onnx::Conv_831[FLOAT, 64x192x1x1]
%onnx::Conv_834[FLOAT, 192x64x1x1]
%onnx::Conv_837[FLOAT, 192x1x5x5]
%onnx::Conv_840[FLOAT, 64x192x1x1]
%onnx::Conv_843[FLOAT, 64x32x1x1]
%onnx::Conv_846[FLOAT, 64x1x3x3]
%onnx::Conv_849[FLOAT, 112x32x1x1]
%onnx::Conv_850[FLOAT, 112]
%onnx::Conv_852[FLOAT, 336x112x1x1]
%onnx::Conv_853[FLOAT, 336]
%onnx::Conv_855[FLOAT, 336x1x5x5]
%onnx::Conv_858[FLOAT, 112x336x1x1]
%onnx::Conv_861[FLOAT, 112x56x1x1]
%onnx::Conv_864[FLOAT, 112x1x5x5]
%onnx::Conv_867[FLOAT, 112x56x1x1]
%onnx::Conv_870[FLOAT, 112x112x1x1]
%onnx::Conv_873[FLOAT, 112x1x3x3]
%onnx::Conv_876[FLOAT, 112x112x1x1]
%onnx::Conv_879[FLOAT, 112x56x1x1]
%onnx::Conv_882[FLOAT, 112x1x5x5]
%onnx::Conv_885[FLOAT, 184x56x1x1]
%onnx::Conv_886[FLOAT, 184]
%onnx::Conv_888[FLOAT, 1104x184x1x1]
%onnx::Conv_889[FLOAT, 1104]
%onnx::Conv_891[FLOAT, 1104x1x3x3]
%onnx::Conv_894[FLOAT, 184x1104x1x1]
%onnx::Conv_897[FLOAT, 184x184x1x1]
%onnx::Conv_900[FLOAT, 184x1x3x3]
%onnx::Conv_903[FLOAT, 184x184x1x1]
%onnx::Conv_906[FLOAT, 184x92x1x1]
%onnx::Conv_909[FLOAT, 184x1x5x5]
%onnx::Conv_912[FLOAT, 352x92x1x1]
%onnx::Conv_913[FLOAT, 352]
%onnx::Conv_915[FLOAT, 1504x352x1x1]
%onnx::Conv_916[FLOAT, 1504]
) {
%onnx::Conv_910 = Identity(%onnx::Conv_886)
%onnx::Conv_907 = Identity(%onnx::Conv_886)
%onnx::Conv_904 = Identity(%onnx::Conv_886)
%onnx::Conv_901 = Identity(%onnx::Conv_886)
%onnx::Conv_898 = Identity(%onnx::Conv_886)
%onnx::Conv_895 = Identity(%onnx::Conv_886)
%onnx::Conv_892 = Identity(%onnx::Conv_889)
%onnx::Conv_883 = Identity(%onnx::Conv_850)
%onnx::Conv_880 = Identity(%onnx::Conv_850)
%onnx::Conv_877 = Identity(%onnx::Conv_850)
%onnx::Conv_874 = Identity(%onnx::Conv_850)
%onnx::Conv_871 = Identity(%onnx::Conv_850)
%onnx::Conv_868 = Identity(%onnx::Conv_850)
%onnx::Conv_865 = Identity(%onnx::Conv_850)
%onnx::Conv_862 = Identity(%onnx::Conv_850)
%onnx::Conv_859 = Identity(%onnx::Conv_850)
%onnx::Conv_856 = Identity(%onnx::Conv_853)
%onnx::Conv_847 = Identity(%onnx::Conv_814)
%onnx::Conv_844 = Identity(%onnx::Conv_814)
%onnx::Conv_841 = Identity(%onnx::Conv_814)
%onnx::Conv_838 = Identity(%onnx::Conv_790)
%onnx::Conv_835 = Identity(%onnx::Conv_790)
%onnx::Conv_832 = Identity(%onnx::Conv_814)
%onnx::Conv_829 = Identity(%onnx::Conv_790)
%onnx::Conv_826 = Identity(%onnx::Conv_790)
%onnx::Conv_823 = Identity(%onnx::Conv_814)
%onnx::Conv_820 = Identity(%onnx::Conv_814)
%onnx::Conv_817 = Identity(%onnx::Conv_814)
%onnx::Conv_811 = Identity(%onnx::Conv_790)
%onnx::Conv_808 = Identity(%onnx::Conv_790)
%onnx::Conv_805 = Identity(%onnx::Conv_778)
%onnx::Conv_802 = Identity(%onnx::Conv_790)
%onnx::Conv_799 = Identity(%onnx::Conv_790)
%onnx::Conv_796 = Identity(%onnx::Conv_778)
%onnx::Conv_793 = Identity(%onnx::Conv_790)
%onnx::Conv_787 = Identity(%onnx::Conv_778)
%onnx::Conv_784 = Identity(%onnx::Conv_778)
%onnx::Conv_781 = Identity(%onnx::Conv_778)
%onnx::Conv_775 = Identity(%onnx::Conv_745)
%onnx::Conv_772 = Identity(%onnx::Conv_745)
%onnx::Conv_769 = Identity(%onnx::Conv_742)
%onnx::Conv_766 = Identity(%onnx::Conv_742)
%onnx::Conv_763 = Identity(%onnx::Conv_742)
%onnx::Conv_760 = Identity(%onnx::Conv_742)
%onnx::Conv_757 = Identity(%onnx::Conv_745)
%onnx::Conv_754 = Identity(%onnx::Conv_745)
%onnx::Conv_751 = Identity(%onnx::Conv_742)
%onnx::Conv_748 = Identity(%onnx::Conv_745)
%onnx::Conv_739 = Identity(%onnx::Conv_724)
%onnx::Conv_736 = Identity(%onnx::Conv_724)
%onnx::Conv_733 = Identity(%onnx::Conv_724)
%onnx::Conv_730 = Identity(%onnx::Conv_727)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_723, %onnx::Conv_724)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_858, %onnx::Conv_859)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_861, %onnx::Conv_862)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_864, %onnx::Conv_865)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_867, %onnx::Conv_868)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_870, %onnx::Conv_871)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_873, %onnx::Conv_874)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_876, %onnx::Conv_877)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_879, %onnx::Conv_880)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_882, %onnx::Conv_883)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_885, %onnx::Conv_886)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_888, %onnx::Conv_889)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_891, %onnx::Conv_892)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_894, %onnx::Conv_895)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_897, %onnx::Conv_898)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_900, %onnx::Conv_901)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_903, %onnx::Conv_904)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_906, %onnx::Conv_907)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_909, %onnx::Conv_910)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_912, %onnx::Conv_913)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_915, %onnx::Conv_916)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%721 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %721
}
|
val_accuracy
| 0
| 78,438,016
| 1,544,092
|
{'zcp_synflow': 76.6471609517443, 'zcp_zen': 68.92240905761719, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.2448970377445221, 'zcp_flops': 78438016.0, 'zcp_grad_norm': 31.82300567626953, 'zcp_grasp': 0.5336532592773438, 'zcp_jacov': -16.052015064699734, 'zcp_l2_norm': 604.9357299804688, 'zcp_nwot': 218.87373100923435, 'zcp_params': 1544092.0, 'zcp_plain': 0.001032708678394556, 'zcp_snip': 57.4855842590332, 'lat_1080ti_1': 0.8128964250060716, 'lat_1080ti_32': 0.8134639055498384, 'lat_1080ti_64': 0.8364009413097413, 'lat_2080ti_1': 0.7822784803171825, 'lat_2080ti_32': 0.8662457771398652, 'lat_2080ti_64': 0.8648727876667688, 'lat_essential_ph_1': 0.3584905660377358, 'lat_eyeriss': 0.6296004254304735, 'lat_fpga': 0.5351428373524376, 'lat_gold_6226': 0.33651329326771084, 'lat_gold_6240': 0.49370995094845416, 'lat_pixel2': 0.32608695652173914, 'lat_pixel3': 0.6570807539695438, 'lat_raspi4': 0.6603350601340195, 'lat_samsung_a50': 0.24210526315789474, 'lat_samsung_s7': 0.1889763779527559, 'lat_silver_4114': 0.5225342495350377, 'lat_silver_4210r': 0.5962993590059015, 'lat_titan_rtx_1': 0.7293166376611931, 'lat_titan_rtx_32': 0.809636563409488, 'lat_titan_rtx_64': 0.8776799646859633, 'lat_titanx_1': 0.38867397053069797, 'lat_titanx_32': 0.8515418698985342, 'lat_titanx_64': 0.7896377958646628, 'lat_titanxp_1': 0.7076924223809021, 'lat_titanxp_32': 0.8490414598163968, 'lat_titanxp_64': 0.8527693079445664}
| |
FBNet_3913
|
FBNet
|
3913
|
3913
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_613[FLOAT, 16x3x3x3]
%onnx::Conv_614[FLOAT, 16]
%onnx::Conv_616[FLOAT, 16x16x1x1]
%onnx::Conv_619[FLOAT, 16x1x5x5]
%onnx::Conv_622[FLOAT, 16x16x1x1]
%onnx::Conv_625[FLOAT, 24x16x1x1]
%onnx::Conv_626[FLOAT, 24]
%onnx::Conv_628[FLOAT, 144x24x1x1]
%onnx::Conv_629[FLOAT, 144]
%onnx::Conv_631[FLOAT, 144x1x5x5]
%onnx::Conv_634[FLOAT, 24x144x1x1]
%onnx::Conv_637[FLOAT, 24x24x1x1]
%onnx::Conv_640[FLOAT, 24x1x3x3]
%onnx::Conv_643[FLOAT, 24x24x1x1]
%onnx::Conv_646[FLOAT, 24x24x1x1]
%onnx::Conv_649[FLOAT, 24x1x5x5]
%onnx::Conv_652[FLOAT, 24x24x1x1]
%onnx::Conv_655[FLOAT, 144x24x1x1]
%onnx::Conv_658[FLOAT, 144x1x3x3]
%onnx::Conv_661[FLOAT, 32x144x1x1]
%onnx::Conv_662[FLOAT, 32]
%onnx::Conv_664[FLOAT, 96x32x1x1]
%onnx::Conv_665[FLOAT, 96]
%onnx::Conv_667[FLOAT, 96x1x5x5]
%onnx::Conv_670[FLOAT, 32x96x1x1]
%onnx::Conv_673[FLOAT, 96x32x1x1]
%onnx::Conv_676[FLOAT, 96x1x5x5]
%onnx::Conv_679[FLOAT, 32x96x1x1]
%onnx::Conv_682[FLOAT, 192x32x1x1]
%onnx::Conv_683[FLOAT, 192]
%onnx::Conv_685[FLOAT, 192x1x3x3]
%onnx::Conv_688[FLOAT, 64x192x1x1]
%onnx::Conv_689[FLOAT, 64]
%onnx::Conv_691[FLOAT, 64x64x1x1]
%onnx::Conv_694[FLOAT, 64x1x3x3]
%onnx::Conv_697[FLOAT, 64x64x1x1]
%onnx::Conv_700[FLOAT, 384x64x1x1]
%onnx::Conv_701[FLOAT, 384]
%onnx::Conv_703[FLOAT, 384x1x5x5]
%onnx::Conv_706[FLOAT, 64x384x1x1]
%onnx::Conv_709[FLOAT, 64x32x1x1]
%onnx::Conv_712[FLOAT, 64x1x5x5]
%onnx::Conv_715[FLOAT, 64x32x1x1]
%onnx::Conv_718[FLOAT, 64x64x1x1]
%onnx::Conv_721[FLOAT, 64x1x3x3]
%onnx::Conv_724[FLOAT, 112x64x1x1]
%onnx::Conv_725[FLOAT, 112]
%onnx::Conv_727[FLOAT, 336x112x1x1]
%onnx::Conv_728[FLOAT, 336]
%onnx::Conv_730[FLOAT, 336x1x5x5]
%onnx::Conv_733[FLOAT, 112x336x1x1]
%onnx::Conv_736[FLOAT, 336x112x1x1]
%onnx::Conv_739[FLOAT, 336x1x3x3]
%onnx::Conv_742[FLOAT, 112x336x1x1]
%onnx::Conv_745[FLOAT, 672x112x1x1]
%onnx::Conv_746[FLOAT, 672]
%onnx::Conv_748[FLOAT, 672x1x5x5]
%onnx::Conv_751[FLOAT, 112x672x1x1]
%onnx::Conv_754[FLOAT, 672x112x1x1]
%onnx::Conv_757[FLOAT, 672x1x5x5]
%onnx::Conv_760[FLOAT, 184x672x1x1]
%onnx::Conv_761[FLOAT, 184]
%onnx::Conv_763[FLOAT, 552x184x1x1]
%onnx::Conv_764[FLOAT, 552]
%onnx::Conv_766[FLOAT, 552x1x3x3]
%onnx::Conv_769[FLOAT, 184x552x1x1]
%onnx::Conv_772[FLOAT, 184x92x1x1]
%onnx::Conv_775[FLOAT, 184x1x3x3]
%onnx::Conv_778[FLOAT, 184x92x1x1]
%onnx::Conv_781[FLOAT, 552x184x1x1]
%onnx::Conv_784[FLOAT, 552x1x3x3]
%onnx::Conv_787[FLOAT, 184x552x1x1]
%onnx::Conv_790[FLOAT, 184x184x1x1]
%onnx::Conv_793[FLOAT, 184x1x3x3]
%onnx::Conv_796[FLOAT, 352x184x1x1]
%onnx::Conv_797[FLOAT, 352]
%onnx::Conv_799[FLOAT, 1504x352x1x1]
%onnx::Conv_800[FLOAT, 1504]
) {
%onnx::Conv_794 = Identity(%onnx::Conv_761)
%onnx::Conv_791 = Identity(%onnx::Conv_761)
%onnx::Conv_788 = Identity(%onnx::Conv_761)
%onnx::Conv_785 = Identity(%onnx::Conv_764)
%onnx::Conv_782 = Identity(%onnx::Conv_764)
%onnx::Conv_779 = Identity(%onnx::Conv_761)
%onnx::Conv_776 = Identity(%onnx::Conv_761)
%onnx::Conv_773 = Identity(%onnx::Conv_761)
%onnx::Conv_770 = Identity(%onnx::Conv_761)
%onnx::Conv_767 = Identity(%onnx::Conv_764)
%onnx::Conv_758 = Identity(%onnx::Conv_746)
%onnx::Conv_755 = Identity(%onnx::Conv_746)
%onnx::Conv_752 = Identity(%onnx::Conv_725)
%onnx::Conv_749 = Identity(%onnx::Conv_746)
%onnx::Conv_743 = Identity(%onnx::Conv_725)
%onnx::Conv_740 = Identity(%onnx::Conv_728)
%onnx::Conv_737 = Identity(%onnx::Conv_728)
%onnx::Conv_734 = Identity(%onnx::Conv_725)
%onnx::Conv_731 = Identity(%onnx::Conv_728)
%onnx::Conv_722 = Identity(%onnx::Conv_689)
%onnx::Conv_719 = Identity(%onnx::Conv_689)
%onnx::Conv_716 = Identity(%onnx::Conv_689)
%onnx::Conv_713 = Identity(%onnx::Conv_689)
%onnx::Conv_710 = Identity(%onnx::Conv_689)
%onnx::Conv_707 = Identity(%onnx::Conv_689)
%onnx::Conv_704 = Identity(%onnx::Conv_701)
%onnx::Conv_698 = Identity(%onnx::Conv_689)
%onnx::Conv_695 = Identity(%onnx::Conv_689)
%onnx::Conv_692 = Identity(%onnx::Conv_689)
%onnx::Conv_686 = Identity(%onnx::Conv_683)
%onnx::Conv_680 = Identity(%onnx::Conv_662)
%onnx::Conv_677 = Identity(%onnx::Conv_665)
%onnx::Conv_674 = Identity(%onnx::Conv_665)
%onnx::Conv_671 = Identity(%onnx::Conv_662)
%onnx::Conv_668 = Identity(%onnx::Conv_665)
%onnx::Conv_659 = Identity(%onnx::Conv_629)
%onnx::Conv_656 = Identity(%onnx::Conv_629)
%onnx::Conv_653 = Identity(%onnx::Conv_626)
%onnx::Conv_650 = Identity(%onnx::Conv_626)
%onnx::Conv_647 = Identity(%onnx::Conv_626)
%onnx::Conv_644 = Identity(%onnx::Conv_626)
%onnx::Conv_641 = Identity(%onnx::Conv_626)
%onnx::Conv_638 = Identity(%onnx::Conv_626)
%onnx::Conv_635 = Identity(%onnx::Conv_626)
%onnx::Conv_632 = Identity(%onnx::Conv_629)
%onnx::Conv_623 = Identity(%onnx::Conv_614)
%onnx::Conv_620 = Identity(%onnx::Conv_614)
%onnx::Conv_617 = Identity(%onnx::Conv_614)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_613, %onnx::Conv_614)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_616, %onnx::Conv_617)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_619, %onnx::Conv_620)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_622, %onnx::Conv_623)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_625, %onnx::Conv_626)
%/cells.1/relu/Relu_output_0 = Relu(%/cells.1/conv/Conv_output_0)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/relu/Relu_output_0, %onnx::Conv_628, %onnx::Conv_629)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_631, %onnx::Conv_632)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_634, %onnx::Conv_635)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/relu/Relu_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_637, %onnx::Conv_638)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_640, %onnx::Conv_641)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_643, %onnx::Conv_644)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_646, %onnx::Conv_647)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_649, %onnx::Conv_650)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_652, %onnx::Conv_653)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_655, %onnx::Conv_656)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_658, %onnx::Conv_659)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%611 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %611
}
|
val_accuracy
| 0
| 81,933,184
| 1,950,164
|
{'zcp_synflow': 82.20978322993385, 'zcp_zen': 72.05484771728516, 'zcp_epe_nas': 9.91617065008695, 'zcp_fisher': 0.17930781841278076, 'zcp_flops': 81933184.0, 'zcp_grad_norm': 26.835308074951172, 'zcp_grasp': 0.048816680908203125, 'zcp_jacov': -16.06717951422153, 'zcp_l2_norm': 676.2997436523438, 'zcp_nwot': 214.31896286107474, 'zcp_params': 1950164.0, 'zcp_plain': -0.0029977206140756607, 'zcp_snip': 46.80640411376953, 'lat_1080ti_1': 0.5550642905951205, 'lat_1080ti_32': 0.5557595504554151, 'lat_1080ti_64': 0.49520112629360696, 'lat_2080ti_1': 0.7310246365287674, 'lat_2080ti_32': 0.6007135152343462, 'lat_2080ti_64': 0.49647958460955854, 'lat_essential_ph_1': 0.37735849056603776, 'lat_eyeriss': 0.5409293455338967, 'lat_fpga': 0.5517865801165656, 'lat_gold_6226': 0.44334087341038037, 'lat_gold_6240': 0.5824605591456314, 'lat_pixel2': 0.3695652173913043, 'lat_pixel3': 0.565634997281796, 'lat_raspi4': 0.5363200468268382, 'lat_samsung_a50': 0.24210526315789474, 'lat_samsung_s7': 0.5196850393700787, 'lat_silver_4114': 0.5765044661378042, 'lat_silver_4210r': 0.657279206697225, 'lat_titan_rtx_1': 0.5742919255461888, 'lat_titan_rtx_32': 0.5630803168645406, 'lat_titan_rtx_64': 0.5115730746372436, 'lat_titanx_1': 0.31011348390832755, 'lat_titanx_32': 0.5339614109851566, 'lat_titanx_64': 0.4838177627349535, 'lat_titanxp_1': 0.5645146941050446, 'lat_titanxp_32': 0.5416410847551639, 'lat_titanxp_64': 0.4999028499413995}
| |
FBNet_2738
|
FBNet
|
2738
|
2738
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_641[FLOAT, 16x3x3x3]
%onnx::Conv_642[FLOAT, 16]
%onnx::Conv_644[FLOAT, 24x16x1x1]
%onnx::Conv_645[FLOAT, 24]
%onnx::Conv_647[FLOAT, 144x24x1x1]
%onnx::Conv_648[FLOAT, 144]
%onnx::Conv_650[FLOAT, 144x1x3x3]
%onnx::Conv_653[FLOAT, 24x144x1x1]
%onnx::Conv_656[FLOAT, 72x24x1x1]
%onnx::Conv_657[FLOAT, 72]
%onnx::Conv_659[FLOAT, 72x1x5x5]
%onnx::Conv_662[FLOAT, 24x72x1x1]
%onnx::Conv_665[FLOAT, 24x12x1x1]
%onnx::Conv_668[FLOAT, 24x1x3x3]
%onnx::Conv_671[FLOAT, 24x12x1x1]
%onnx::Conv_674[FLOAT, 24x12x1x1]
%onnx::Conv_677[FLOAT, 24x1x3x3]
%onnx::Conv_680[FLOAT, 32x12x1x1]
%onnx::Conv_681[FLOAT, 32]
%onnx::Conv_683[FLOAT, 32x32x1x1]
%onnx::Conv_686[FLOAT, 32x1x3x3]
%onnx::Conv_689[FLOAT, 32x32x1x1]
%onnx::Conv_692[FLOAT, 32x16x1x1]
%onnx::Conv_695[FLOAT, 32x1x5x5]
%onnx::Conv_698[FLOAT, 32x16x1x1]
%onnx::Conv_701[FLOAT, 192x32x1x1]
%onnx::Conv_702[FLOAT, 192]
%onnx::Conv_704[FLOAT, 192x1x3x3]
%onnx::Conv_707[FLOAT, 32x192x1x1]
%onnx::Conv_710[FLOAT, 192x32x1x1]
%onnx::Conv_713[FLOAT, 192x1x3x3]
%onnx::Conv_716[FLOAT, 64x192x1x1]
%onnx::Conv_717[FLOAT, 64]
%onnx::Conv_719[FLOAT, 64x32x1x1]
%onnx::Conv_722[FLOAT, 64x1x3x3]
%onnx::Conv_725[FLOAT, 64x32x1x1]
%onnx::Conv_728[FLOAT, 64x64x1x1]
%onnx::Conv_731[FLOAT, 64x1x3x3]
%onnx::Conv_734[FLOAT, 64x64x1x1]
%onnx::Conv_737[FLOAT, 64x64x1x1]
%onnx::Conv_740[FLOAT, 64x1x5x5]
%onnx::Conv_743[FLOAT, 64x64x1x1]
%onnx::Conv_746[FLOAT, 64x32x1x1]
%onnx::Conv_749[FLOAT, 64x1x5x5]
%onnx::Conv_752[FLOAT, 112x32x1x1]
%onnx::Conv_753[FLOAT, 112]
%onnx::Conv_755[FLOAT, 112x112x1x1]
%onnx::Conv_758[FLOAT, 112x1x3x3]
%onnx::Conv_761[FLOAT, 112x112x1x1]
%onnx::Conv_764[FLOAT, 112x112x1x1]
%onnx::Conv_767[FLOAT, 112x1x5x5]
%onnx::Conv_770[FLOAT, 112x112x1x1]
%onnx::Conv_773[FLOAT, 672x112x1x1]
%onnx::Conv_774[FLOAT, 672]
%onnx::Conv_776[FLOAT, 672x1x3x3]
%onnx::Conv_779[FLOAT, 112x672x1x1]
%onnx::Conv_782[FLOAT, 336x112x1x1]
%onnx::Conv_783[FLOAT, 336]
%onnx::Conv_785[FLOAT, 336x1x3x3]
%onnx::Conv_788[FLOAT, 184x336x1x1]
%onnx::Conv_789[FLOAT, 184]
%onnx::Conv_791[FLOAT, 1104x184x1x1]
%onnx::Conv_792[FLOAT, 1104]
%onnx::Conv_794[FLOAT, 1104x1x3x3]
%onnx::Conv_797[FLOAT, 184x1104x1x1]
%onnx::Conv_800[FLOAT, 1104x184x1x1]
%onnx::Conv_803[FLOAT, 1104x1x3x3]
%onnx::Conv_806[FLOAT, 184x1104x1x1]
%onnx::Conv_809[FLOAT, 552x184x1x1]
%onnx::Conv_810[FLOAT, 552]
%onnx::Conv_812[FLOAT, 552x1x5x5]
%onnx::Conv_815[FLOAT, 352x552x1x1]
%onnx::Conv_816[FLOAT, 352]
%onnx::Conv_818[FLOAT, 1504x352x1x1]
%onnx::Conv_819[FLOAT, 1504]
) {
%onnx::Conv_813 = Identity(%onnx::Conv_810)
%onnx::Conv_807 = Identity(%onnx::Conv_789)
%onnx::Conv_804 = Identity(%onnx::Conv_792)
%onnx::Conv_801 = Identity(%onnx::Conv_792)
%onnx::Conv_798 = Identity(%onnx::Conv_789)
%onnx::Conv_795 = Identity(%onnx::Conv_792)
%onnx::Conv_786 = Identity(%onnx::Conv_783)
%onnx::Conv_780 = Identity(%onnx::Conv_753)
%onnx::Conv_777 = Identity(%onnx::Conv_774)
%onnx::Conv_771 = Identity(%onnx::Conv_753)
%onnx::Conv_768 = Identity(%onnx::Conv_753)
%onnx::Conv_765 = Identity(%onnx::Conv_753)
%onnx::Conv_762 = Identity(%onnx::Conv_753)
%onnx::Conv_759 = Identity(%onnx::Conv_753)
%onnx::Conv_756 = Identity(%onnx::Conv_753)
%onnx::Conv_750 = Identity(%onnx::Conv_717)
%onnx::Conv_747 = Identity(%onnx::Conv_717)
%onnx::Conv_744 = Identity(%onnx::Conv_717)
%onnx::Conv_741 = Identity(%onnx::Conv_717)
%onnx::Conv_738 = Identity(%onnx::Conv_717)
%onnx::Conv_735 = Identity(%onnx::Conv_717)
%onnx::Conv_732 = Identity(%onnx::Conv_717)
%onnx::Conv_729 = Identity(%onnx::Conv_717)
%onnx::Conv_726 = Identity(%onnx::Conv_717)
%onnx::Conv_723 = Identity(%onnx::Conv_717)
%onnx::Conv_720 = Identity(%onnx::Conv_717)
%onnx::Conv_714 = Identity(%onnx::Conv_702)
%onnx::Conv_711 = Identity(%onnx::Conv_702)
%onnx::Conv_708 = Identity(%onnx::Conv_681)
%onnx::Conv_705 = Identity(%onnx::Conv_702)
%onnx::Conv_699 = Identity(%onnx::Conv_681)
%onnx::Conv_696 = Identity(%onnx::Conv_681)
%onnx::Conv_693 = Identity(%onnx::Conv_681)
%onnx::Conv_690 = Identity(%onnx::Conv_681)
%onnx::Conv_687 = Identity(%onnx::Conv_681)
%onnx::Conv_684 = Identity(%onnx::Conv_681)
%onnx::Conv_678 = Identity(%onnx::Conv_645)
%onnx::Conv_675 = Identity(%onnx::Conv_645)
%onnx::Conv_672 = Identity(%onnx::Conv_645)
%onnx::Conv_669 = Identity(%onnx::Conv_645)
%onnx::Conv_666 = Identity(%onnx::Conv_645)
%onnx::Conv_663 = Identity(%onnx::Conv_645)
%onnx::Conv_660 = Identity(%onnx::Conv_657)
%onnx::Conv_654 = Identity(%onnx::Conv_645)
%onnx::Conv_651 = Identity(%onnx::Conv_648)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_641, %onnx::Conv_642)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.1/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.1/relu/Relu_output_0 = Relu(%/cells.1/conv/Conv_output_0)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/relu/Relu_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/relu/Relu_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_803, %onnx::Conv_804)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_806, %onnx::Conv_807)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_809, %onnx::Conv_810)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_812, %onnx::Conv_813)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_815, %onnx::Conv_816)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_818, %onnx::Conv_819)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%639 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %639
}
|
val_accuracy
| 0
| 71,730,048
| 2,246,340
|
{'zcp_synflow': 72.84851378117553, 'zcp_zen': 65.2976303100586, 'zcp_epe_nas': 6.983261395872902, 'zcp_fisher': 0.12805958092212677, 'zcp_flops': 71730048.0, 'zcp_grad_norm': 21.338279724121094, 'zcp_grasp': -0.07602691650390625, 'zcp_jacov': -16.05470665270323, 'zcp_l2_norm': 622.41552734375, 'zcp_nwot': 212.2345303486374, 'zcp_params': 2246340.0, 'zcp_plain': 0.0013355532428249717, 'zcp_snip': 39.38079071044922, 'lat_1080ti_1': 0.5458260476060299, 'lat_1080ti_32': 0.5552843677869803, 'lat_1080ti_64': 0.3886870518452004, 'lat_2080ti_1': 0.554330531314761, 'lat_2080ti_32': 0.556740037199859, 'lat_2080ti_64': 0.44091693275011334, 'lat_essential_ph_1': 0.24528301886792453, 'lat_eyeriss': 0.45621917004382645, 'lat_fpga': 0.5330945289385408, 'lat_gold_6226': 0.4043197004129737, 'lat_gold_6240': 0.5643288535739407, 'lat_pixel2': 0.32608695652173914, 'lat_pixel3': 0.4256228027715719, 'lat_raspi4': 0.5236110153017627, 'lat_samsung_a50': 0.22105263157894736, 'lat_samsung_s7': 0.18110236220472442, 'lat_silver_4114': 0.679476727233916, 'lat_silver_4210r': 0.5654605955354964, 'lat_titan_rtx_1': 0.5303348246873949, 'lat_titan_rtx_32': 0.5288039953942021, 'lat_titan_rtx_64': 0.45532385261421116, 'lat_titanx_1': 0.2927960253336996, 'lat_titanx_32': 0.4809621413620679, 'lat_titanx_64': 0.38392105243898345, 'lat_titanxp_1': 0.5147352077143934, 'lat_titanxp_32': 0.4920253397108855, 'lat_titanxp_64': 0.4156876656501533}
| |
FBNet_503
|
FBNet
|
503
|
503
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_659[FLOAT, 16x3x3x3]
%onnx::Conv_660[FLOAT, 16]
%onnx::Conv_662[FLOAT, 48x16x1x1]
%onnx::Conv_663[FLOAT, 48]
%onnx::Conv_665[FLOAT, 48x1x3x3]
%onnx::Conv_668[FLOAT, 16x48x1x1]
%onnx::Conv_671[FLOAT, 16x16x1x1]
%onnx::Conv_674[FLOAT, 16x1x3x3]
%onnx::Conv_677[FLOAT, 24x16x1x1]
%onnx::Conv_678[FLOAT, 24]
%onnx::Conv_680[FLOAT, 24x24x1x1]
%onnx::Conv_683[FLOAT, 24x1x5x5]
%onnx::Conv_686[FLOAT, 24x24x1x1]
%onnx::Conv_689[FLOAT, 24x24x1x1]
%onnx::Conv_692[FLOAT, 24x1x5x5]
%onnx::Conv_695[FLOAT, 24x24x1x1]
%onnx::Conv_698[FLOAT, 72x24x1x1]
%onnx::Conv_699[FLOAT, 72]
%onnx::Conv_701[FLOAT, 72x1x3x3]
%onnx::Conv_704[FLOAT, 24x72x1x1]
%onnx::Conv_707[FLOAT, 144x24x1x1]
%onnx::Conv_708[FLOAT, 144]
%onnx::Conv_710[FLOAT, 144x1x5x5]
%onnx::Conv_713[FLOAT, 32x144x1x1]
%onnx::Conv_714[FLOAT, 32]
%onnx::Conv_716[FLOAT, 32x16x1x1]
%onnx::Conv_719[FLOAT, 32x1x5x5]
%onnx::Conv_722[FLOAT, 32x16x1x1]
%onnx::Conv_725[FLOAT, 32x32x1x1]
%onnx::Conv_728[FLOAT, 32x1x3x3]
%onnx::Conv_731[FLOAT, 32x32x1x1]
%onnx::Conv_734[FLOAT, 192x32x1x1]
%onnx::Conv_735[FLOAT, 192]
%onnx::Conv_737[FLOAT, 192x1x3x3]
%onnx::Conv_740[FLOAT, 32x192x1x1]
%onnx::Conv_743[FLOAT, 96x32x1x1]
%onnx::Conv_744[FLOAT, 96]
%onnx::Conv_746[FLOAT, 96x1x3x3]
%onnx::Conv_749[FLOAT, 64x96x1x1]
%onnx::Conv_750[FLOAT, 64]
%onnx::Conv_752[FLOAT, 64x32x1x1]
%onnx::Conv_755[FLOAT, 64x1x3x3]
%onnx::Conv_758[FLOAT, 64x32x1x1]
%onnx::Conv_761[FLOAT, 384x64x1x1]
%onnx::Conv_762[FLOAT, 384]
%onnx::Conv_764[FLOAT, 384x1x3x3]
%onnx::Conv_767[FLOAT, 64x384x1x1]
%onnx::Conv_770[FLOAT, 384x64x1x1]
%onnx::Conv_773[FLOAT, 384x1x3x3]
%onnx::Conv_776[FLOAT, 64x384x1x1]
%onnx::Conv_779[FLOAT, 192x64x1x1]
%onnx::Conv_782[FLOAT, 192x1x3x3]
%onnx::Conv_785[FLOAT, 112x192x1x1]
%onnx::Conv_786[FLOAT, 112]
%onnx::Conv_788[FLOAT, 112x112x1x1]
%onnx::Conv_791[FLOAT, 112x1x3x3]
%onnx::Conv_794[FLOAT, 112x112x1x1]
%onnx::Conv_797[FLOAT, 184x112x1x1]
%onnx::Conv_798[FLOAT, 184]
%onnx::Conv_800[FLOAT, 184x92x1x1]
%onnx::Conv_803[FLOAT, 184x1x3x3]
%onnx::Conv_806[FLOAT, 184x92x1x1]
%onnx::Conv_809[FLOAT, 184x92x1x1]
%onnx::Conv_812[FLOAT, 184x1x3x3]
%onnx::Conv_815[FLOAT, 184x92x1x1]
%onnx::Conv_818[FLOAT, 184x92x1x1]
%onnx::Conv_821[FLOAT, 184x1x5x5]
%onnx::Conv_824[FLOAT, 184x92x1x1]
%onnx::Conv_827[FLOAT, 184x92x1x1]
%onnx::Conv_830[FLOAT, 184x1x3x3]
%onnx::Conv_833[FLOAT, 352x92x1x1]
%onnx::Conv_834[FLOAT, 352]
%onnx::Conv_836[FLOAT, 1504x352x1x1]
%onnx::Conv_837[FLOAT, 1504]
) {
%onnx::Conv_831 = Identity(%onnx::Conv_798)
%onnx::Conv_828 = Identity(%onnx::Conv_798)
%onnx::Conv_825 = Identity(%onnx::Conv_798)
%onnx::Conv_822 = Identity(%onnx::Conv_798)
%onnx::Conv_819 = Identity(%onnx::Conv_798)
%onnx::Conv_816 = Identity(%onnx::Conv_798)
%onnx::Conv_813 = Identity(%onnx::Conv_798)
%onnx::Conv_810 = Identity(%onnx::Conv_798)
%onnx::Conv_807 = Identity(%onnx::Conv_798)
%onnx::Conv_804 = Identity(%onnx::Conv_798)
%onnx::Conv_801 = Identity(%onnx::Conv_798)
%onnx::Conv_795 = Identity(%onnx::Conv_786)
%onnx::Conv_792 = Identity(%onnx::Conv_786)
%onnx::Conv_789 = Identity(%onnx::Conv_786)
%onnx::Conv_783 = Identity(%onnx::Conv_735)
%onnx::Conv_780 = Identity(%onnx::Conv_735)
%onnx::Conv_777 = Identity(%onnx::Conv_750)
%onnx::Conv_774 = Identity(%onnx::Conv_762)
%onnx::Conv_771 = Identity(%onnx::Conv_762)
%onnx::Conv_768 = Identity(%onnx::Conv_750)
%onnx::Conv_765 = Identity(%onnx::Conv_762)
%onnx::Conv_759 = Identity(%onnx::Conv_750)
%onnx::Conv_756 = Identity(%onnx::Conv_750)
%onnx::Conv_753 = Identity(%onnx::Conv_750)
%onnx::Conv_747 = Identity(%onnx::Conv_744)
%onnx::Conv_741 = Identity(%onnx::Conv_714)
%onnx::Conv_738 = Identity(%onnx::Conv_735)
%onnx::Conv_732 = Identity(%onnx::Conv_714)
%onnx::Conv_729 = Identity(%onnx::Conv_714)
%onnx::Conv_726 = Identity(%onnx::Conv_714)
%onnx::Conv_723 = Identity(%onnx::Conv_714)
%onnx::Conv_720 = Identity(%onnx::Conv_714)
%onnx::Conv_717 = Identity(%onnx::Conv_714)
%onnx::Conv_711 = Identity(%onnx::Conv_708)
%onnx::Conv_705 = Identity(%onnx::Conv_678)
%onnx::Conv_702 = Identity(%onnx::Conv_699)
%onnx::Conv_696 = Identity(%onnx::Conv_678)
%onnx::Conv_693 = Identity(%onnx::Conv_678)
%onnx::Conv_690 = Identity(%onnx::Conv_678)
%onnx::Conv_687 = Identity(%onnx::Conv_678)
%onnx::Conv_684 = Identity(%onnx::Conv_678)
%onnx::Conv_681 = Identity(%onnx::Conv_678)
%onnx::Conv_675 = Identity(%onnx::Conv_660)
%onnx::Conv_672 = Identity(%onnx::Conv_660)
%onnx::Conv_669 = Identity(%onnx::Conv_660)
%onnx::Conv_666 = Identity(%onnx::Conv_663)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_659, %onnx::Conv_660)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_output_0 = Reshape(%/cells.18/nl/Relu_output_0, %/cells.18/shuffle/Constant_output_0)
%/cells.18/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.18/shuffle/Reshape_output_0)
%/cells.18/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_1_output_0 = Reshape(%/cells.18/shuffle/Transpose_output_0, %/cells.18/shuffle/Constant_1_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/shuffle/Reshape_1_output_0, %onnx::Conv_803, %onnx::Conv_804)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_806, %onnx::Conv_807)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_809, %onnx::Conv_810)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_812, %onnx::Conv_813)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_815, %onnx::Conv_816)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_818, %onnx::Conv_819)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_821, %onnx::Conv_822)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_824, %onnx::Conv_825)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_827, %onnx::Conv_828)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_830, %onnx::Conv_831)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_833, %onnx::Conv_834)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_836, %onnx::Conv_837)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%657 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %657
}
|
val_accuracy
| 0
| 47,778,048
| 1,100,044
|
{'zcp_synflow': 70.76507655613875, 'zcp_zen': 60.03416442871094, 'zcp_epe_nas': 39.7230173458465, 'zcp_fisher': 0.11241824924945831, 'zcp_flops': 47778048.0, 'zcp_grad_norm': 23.10270881652832, 'zcp_grasp': -0.04522418975830078, 'zcp_jacov': -16.076908720824648, 'zcp_l2_norm': 508.8357849121094, 'zcp_nwot': 210.68705490293985, 'zcp_params': 1100044.0, 'zcp_plain': 0.000754979089833796, 'zcp_snip': 35.94034957885742, 'lat_1080ti_1': 0.625024597815043, 'lat_1080ti_32': 0.42835101385024854, 'lat_1080ti_64': 0.33722490212190526, 'lat_2080ti_1': 0.5614394781233161, 'lat_2080ti_32': 0.43211431493625857, 'lat_2080ti_64': 0.37284287746191114, 'lat_essential_ph_1': 0.2641509433962264, 'lat_eyeriss': 0.21099150972805447, 'lat_fpga': 0.17737163439179082, 'lat_gold_6226': 0.1324024166498086, 'lat_gold_6240': 0.3630665321617702, 'lat_pixel2': 0.13043478260869565, 'lat_pixel3': 0.19206582216084295, 'lat_raspi4': 0.21133669875217936, 'lat_samsung_a50': 0.24210526315789474, 'lat_samsung_s7': 0.12598425196850394, 'lat_silver_4114': 0.36984787200143443, 'lat_silver_4210r': 0.35357094498931996, 'lat_titan_rtx_1': 0.539291218108639, 'lat_titan_rtx_32': 0.4409073041190104, 'lat_titan_rtx_64': 0.38276803990788805, 'lat_titanx_1': 0.2806608519348383, 'lat_titanx_32': 0.38372372550506495, 'lat_titanx_64': 0.34474816409328085, 'lat_titanxp_1': 0.5057749382900208, 'lat_titanxp_32': 0.42617994274475185, 'lat_titanxp_64': 0.3686188805200084}
| |
FBNet_4484
|
FBNet
|
4484
|
4484
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_531[FLOAT, 16x3x3x3]
%onnx::Conv_532[FLOAT, 16]
%onnx::Conv_534[FLOAT, 96x16x1x1]
%onnx::Conv_535[FLOAT, 96]
%onnx::Conv_537[FLOAT, 96x1x3x3]
%onnx::Conv_540[FLOAT, 16x96x1x1]
%onnx::Conv_543[FLOAT, 16x16x1x1]
%onnx::Conv_546[FLOAT, 16x1x5x5]
%onnx::Conv_549[FLOAT, 24x16x1x1]
%onnx::Conv_550[FLOAT, 24]
%onnx::Conv_552[FLOAT, 144x24x1x1]
%onnx::Conv_553[FLOAT, 144]
%onnx::Conv_555[FLOAT, 144x1x5x5]
%onnx::Conv_558[FLOAT, 24x144x1x1]
%onnx::Conv_561[FLOAT, 72x24x1x1]
%onnx::Conv_562[FLOAT, 72]
%onnx::Conv_564[FLOAT, 72x1x3x3]
%onnx::Conv_567[FLOAT, 24x72x1x1]
%onnx::Conv_570[FLOAT, 144x24x1x1]
%onnx::Conv_573[FLOAT, 144x1x3x3]
%onnx::Conv_576[FLOAT, 32x144x1x1]
%onnx::Conv_577[FLOAT, 32]
%onnx::Conv_579[FLOAT, 96x32x1x1]
%onnx::Conv_582[FLOAT, 96x1x5x5]
%onnx::Conv_585[FLOAT, 32x96x1x1]
%onnx::Conv_588[FLOAT, 96x32x1x1]
%onnx::Conv_591[FLOAT, 96x1x3x3]
%onnx::Conv_594[FLOAT, 32x96x1x1]
%onnx::Conv_597[FLOAT, 192x32x1x1]
%onnx::Conv_598[FLOAT, 192]
%onnx::Conv_600[FLOAT, 192x1x5x5]
%onnx::Conv_603[FLOAT, 64x192x1x1]
%onnx::Conv_604[FLOAT, 64]
%onnx::Conv_606[FLOAT, 64x64x1x1]
%onnx::Conv_609[FLOAT, 64x1x5x5]
%onnx::Conv_612[FLOAT, 64x64x1x1]
%onnx::Conv_615[FLOAT, 192x64x1x1]
%onnx::Conv_618[FLOAT, 192x1x5x5]
%onnx::Conv_621[FLOAT, 64x192x1x1]
%onnx::Conv_624[FLOAT, 384x64x1x1]
%onnx::Conv_625[FLOAT, 384]
%onnx::Conv_627[FLOAT, 384x1x3x3]
%onnx::Conv_630[FLOAT, 112x384x1x1]
%onnx::Conv_631[FLOAT, 112]
%onnx::Conv_633[FLOAT, 112x56x1x1]
%onnx::Conv_636[FLOAT, 112x1x3x3]
%onnx::Conv_639[FLOAT, 112x56x1x1]
%onnx::Conv_642[FLOAT, 112x112x1x1]
%onnx::Conv_645[FLOAT, 112x1x3x3]
%onnx::Conv_648[FLOAT, 112x112x1x1]
%onnx::Conv_651[FLOAT, 112x56x1x1]
%onnx::Conv_654[FLOAT, 112x1x3x3]
%onnx::Conv_657[FLOAT, 112x56x1x1]
%onnx::Conv_660[FLOAT, 184x112x1x1]
%onnx::Conv_661[FLOAT, 184]
%onnx::Conv_663[FLOAT, 1104x184x1x1]
%onnx::Conv_664[FLOAT, 1104]
%onnx::Conv_666[FLOAT, 1104x1x5x5]
%onnx::Conv_669[FLOAT, 184x1104x1x1]
%onnx::Conv_672[FLOAT, 1104x184x1x1]
%onnx::Conv_675[FLOAT, 1104x1x3x3]
%onnx::Conv_678[FLOAT, 184x1104x1x1]
%onnx::Conv_681[FLOAT, 184x184x1x1]
%onnx::Conv_684[FLOAT, 184x1x5x5]
%onnx::Conv_687[FLOAT, 352x184x1x1]
%onnx::Conv_688[FLOAT, 352]
%onnx::Conv_690[FLOAT, 1504x352x1x1]
%onnx::Conv_691[FLOAT, 1504]
) {
%onnx::Conv_685 = Identity(%onnx::Conv_661)
%onnx::Conv_682 = Identity(%onnx::Conv_661)
%onnx::Conv_679 = Identity(%onnx::Conv_661)
%onnx::Conv_676 = Identity(%onnx::Conv_664)
%onnx::Conv_673 = Identity(%onnx::Conv_664)
%onnx::Conv_670 = Identity(%onnx::Conv_661)
%onnx::Conv_667 = Identity(%onnx::Conv_664)
%onnx::Conv_658 = Identity(%onnx::Conv_631)
%onnx::Conv_655 = Identity(%onnx::Conv_631)
%onnx::Conv_652 = Identity(%onnx::Conv_631)
%onnx::Conv_649 = Identity(%onnx::Conv_631)
%onnx::Conv_646 = Identity(%onnx::Conv_631)
%onnx::Conv_643 = Identity(%onnx::Conv_631)
%onnx::Conv_640 = Identity(%onnx::Conv_631)
%onnx::Conv_637 = Identity(%onnx::Conv_631)
%onnx::Conv_634 = Identity(%onnx::Conv_631)
%onnx::Conv_628 = Identity(%onnx::Conv_625)
%onnx::Conv_622 = Identity(%onnx::Conv_604)
%onnx::Conv_619 = Identity(%onnx::Conv_598)
%onnx::Conv_616 = Identity(%onnx::Conv_598)
%onnx::Conv_613 = Identity(%onnx::Conv_604)
%onnx::Conv_610 = Identity(%onnx::Conv_604)
%onnx::Conv_607 = Identity(%onnx::Conv_604)
%onnx::Conv_601 = Identity(%onnx::Conv_598)
%onnx::Conv_595 = Identity(%onnx::Conv_577)
%onnx::Conv_592 = Identity(%onnx::Conv_535)
%onnx::Conv_589 = Identity(%onnx::Conv_535)
%onnx::Conv_586 = Identity(%onnx::Conv_577)
%onnx::Conv_583 = Identity(%onnx::Conv_535)
%onnx::Conv_580 = Identity(%onnx::Conv_535)
%onnx::Conv_574 = Identity(%onnx::Conv_553)
%onnx::Conv_571 = Identity(%onnx::Conv_553)
%onnx::Conv_568 = Identity(%onnx::Conv_550)
%onnx::Conv_565 = Identity(%onnx::Conv_562)
%onnx::Conv_559 = Identity(%onnx::Conv_550)
%onnx::Conv_556 = Identity(%onnx::Conv_553)
%onnx::Conv_547 = Identity(%onnx::Conv_532)
%onnx::Conv_544 = Identity(%onnx::Conv_532)
%onnx::Conv_541 = Identity(%onnx::Conv_532)
%onnx::Conv_538 = Identity(%onnx::Conv_535)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_531, %onnx::Conv_532)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_534, %onnx::Conv_535)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_537, %onnx::Conv_538)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_540, %onnx::Conv_541)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_543, %onnx::Conv_544)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_546, %onnx::Conv_547)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_549, %onnx::Conv_550)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_552, %onnx::Conv_553)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_555, %onnx::Conv_556)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_558, %onnx::Conv_559)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_561, %onnx::Conv_562)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_564, %onnx::Conv_565)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_567, %onnx::Conv_568)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_570, %onnx::Conv_571)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_573, %onnx::Conv_574)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_576, %onnx::Conv_577)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_579, %onnx::Conv_580)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_582, %onnx::Conv_583)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_585, %onnx::Conv_586)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_588, %onnx::Conv_589)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_591, %onnx::Conv_592)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_594, %onnx::Conv_595)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_597, %onnx::Conv_598)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_600, %onnx::Conv_601)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_603, %onnx::Conv_604)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_606, %onnx::Conv_607)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_609, %onnx::Conv_610)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_612, %onnx::Conv_613)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_615, %onnx::Conv_616)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_618, %onnx::Conv_619)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_621, %onnx::Conv_622)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_624, %onnx::Conv_625)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_627, %onnx::Conv_628)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_630, %onnx::Conv_631)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%529 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %529
}
|
val_accuracy
| 0
| 70,322,432
| 1,908,788
|
{'zcp_synflow': 70.73774720207693, 'zcp_zen': 61.20918273925781, 'zcp_epe_nas': 20.33681342171534, 'zcp_fisher': 0.07621387392282486, 'zcp_flops': 70322432.0, 'zcp_grad_norm': 21.387788772583008, 'zcp_grasp': -0.28733253479003906, 'zcp_jacov': -16.068720401975668, 'zcp_l2_norm': 573.1365966796875, 'zcp_nwot': 215.2055751285378, 'zcp_params': 1908788.0, 'zcp_plain': -0.0068490393459796906, 'zcp_snip': 37.17293930053711, 'lat_1080ti_1': 0.334345748803838, 'lat_1080ti_32': 0.39305672240592165, 'lat_1080ti_64': 0.45300191938106826, 'lat_2080ti_1': 0.316357883242826, 'lat_2080ti_32': 0.38658655567358396, 'lat_2080ti_64': 0.43898730283519916, 'lat_essential_ph_1': 0.22641509433962265, 'lat_eyeriss': 0.5087103221902336, 'lat_fpga': 0.4621557703915533, 'lat_gold_6226': 0.3513516396057954, 'lat_gold_6240': 0.3083092187054644, 'lat_pixel2': 0.41304347826086957, 'lat_pixel3': 0.46790204586191947, 'lat_raspi4': 0.507890508533322, 'lat_samsung_a50': 0.29473684210526313, 'lat_samsung_s7': 0.15748031496062992, 'lat_silver_4114': 0.33506650385328496, 'lat_silver_4210r': 0.3081645178604927, 'lat_titan_rtx_1': 0.2850407605844965, 'lat_titan_rtx_32': 0.32969449644576343, 'lat_titan_rtx_64': 0.4100346849509639, 'lat_titanx_1': 0.14845142221422622, 'lat_titanx_32': 0.3648865762402829, 'lat_titanx_64': 0.422685553311064, 'lat_titanxp_1': 0.2791378271231885, 'lat_titanxp_32': 0.3632748873786759, 'lat_titanxp_64': 0.43050809202385437}
| |
FBNet_4562
|
FBNet
|
4562
|
4562
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_533[FLOAT, 16x3x3x3]
%onnx::Conv_534[FLOAT, 16]
%onnx::Conv_536[FLOAT, 16x8x1x1]
%onnx::Conv_539[FLOAT, 16x1x5x5]
%onnx::Conv_542[FLOAT, 16x8x1x1]
%onnx::Conv_545[FLOAT, 16x8x1x1]
%onnx::Conv_548[FLOAT, 16x1x3x3]
%onnx::Conv_551[FLOAT, 24x8x1x1]
%onnx::Conv_552[FLOAT, 24]
%onnx::Conv_554[FLOAT, 24x12x1x1]
%onnx::Conv_557[FLOAT, 24x1x3x3]
%onnx::Conv_560[FLOAT, 24x12x1x1]
%onnx::Conv_563[FLOAT, 24x24x1x1]
%onnx::Conv_566[FLOAT, 24x1x3x3]
%onnx::Conv_569[FLOAT, 24x24x1x1]
%onnx::Conv_572[FLOAT, 24x24x1x1]
%onnx::Conv_575[FLOAT, 24x1x3x3]
%onnx::Conv_578[FLOAT, 32x24x1x1]
%onnx::Conv_579[FLOAT, 32]
%onnx::Conv_581[FLOAT, 32x32x1x1]
%onnx::Conv_584[FLOAT, 32x1x3x3]
%onnx::Conv_587[FLOAT, 32x32x1x1]
%onnx::Conv_590[FLOAT, 32x32x1x1]
%onnx::Conv_593[FLOAT, 32x1x3x3]
%onnx::Conv_596[FLOAT, 32x32x1x1]
%onnx::Conv_599[FLOAT, 64x32x1x1]
%onnx::Conv_600[FLOAT, 64]
%onnx::Conv_602[FLOAT, 64x64x1x1]
%onnx::Conv_605[FLOAT, 64x1x5x5]
%onnx::Conv_608[FLOAT, 64x64x1x1]
%onnx::Conv_611[FLOAT, 112x64x1x1]
%onnx::Conv_612[FLOAT, 112]
%onnx::Conv_614[FLOAT, 336x112x1x1]
%onnx::Conv_615[FLOAT, 336]
%onnx::Conv_617[FLOAT, 336x1x3x3]
%onnx::Conv_620[FLOAT, 112x336x1x1]
%onnx::Conv_623[FLOAT, 672x112x1x1]
%onnx::Conv_624[FLOAT, 672]
%onnx::Conv_626[FLOAT, 672x1x5x5]
%onnx::Conv_629[FLOAT, 112x672x1x1]
%onnx::Conv_632[FLOAT, 672x112x1x1]
%onnx::Conv_635[FLOAT, 672x1x3x3]
%onnx::Conv_638[FLOAT, 112x672x1x1]
%onnx::Conv_641[FLOAT, 336x112x1x1]
%onnx::Conv_644[FLOAT, 336x1x5x5]
%onnx::Conv_647[FLOAT, 184x336x1x1]
%onnx::Conv_648[FLOAT, 184]
%onnx::Conv_650[FLOAT, 552x184x1x1]
%onnx::Conv_651[FLOAT, 552]
%onnx::Conv_653[FLOAT, 552x1x3x3]
%onnx::Conv_656[FLOAT, 184x552x1x1]
%onnx::Conv_659[FLOAT, 184x184x1x1]
%onnx::Conv_662[FLOAT, 184x1x3x3]
%onnx::Conv_665[FLOAT, 184x184x1x1]
%onnx::Conv_668[FLOAT, 1104x184x1x1]
%onnx::Conv_669[FLOAT, 1104]
%onnx::Conv_671[FLOAT, 1104x1x3x3]
%onnx::Conv_674[FLOAT, 184x1104x1x1]
%onnx::Conv_677[FLOAT, 552x184x1x1]
%onnx::Conv_680[FLOAT, 552x1x3x3]
%onnx::Conv_683[FLOAT, 352x552x1x1]
%onnx::Conv_684[FLOAT, 352]
%onnx::Conv_686[FLOAT, 1504x352x1x1]
%onnx::Conv_687[FLOAT, 1504]
) {
%onnx::Conv_681 = Identity(%onnx::Conv_651)
%onnx::Conv_678 = Identity(%onnx::Conv_651)
%onnx::Conv_675 = Identity(%onnx::Conv_648)
%onnx::Conv_672 = Identity(%onnx::Conv_669)
%onnx::Conv_666 = Identity(%onnx::Conv_648)
%onnx::Conv_663 = Identity(%onnx::Conv_648)
%onnx::Conv_660 = Identity(%onnx::Conv_648)
%onnx::Conv_657 = Identity(%onnx::Conv_648)
%onnx::Conv_654 = Identity(%onnx::Conv_651)
%onnx::Conv_645 = Identity(%onnx::Conv_615)
%onnx::Conv_642 = Identity(%onnx::Conv_615)
%onnx::Conv_639 = Identity(%onnx::Conv_612)
%onnx::Conv_636 = Identity(%onnx::Conv_624)
%onnx::Conv_633 = Identity(%onnx::Conv_624)
%onnx::Conv_630 = Identity(%onnx::Conv_612)
%onnx::Conv_627 = Identity(%onnx::Conv_624)
%onnx::Conv_621 = Identity(%onnx::Conv_612)
%onnx::Conv_618 = Identity(%onnx::Conv_615)
%onnx::Conv_609 = Identity(%onnx::Conv_600)
%onnx::Conv_606 = Identity(%onnx::Conv_600)
%onnx::Conv_603 = Identity(%onnx::Conv_600)
%onnx::Conv_597 = Identity(%onnx::Conv_579)
%onnx::Conv_594 = Identity(%onnx::Conv_579)
%onnx::Conv_591 = Identity(%onnx::Conv_579)
%onnx::Conv_588 = Identity(%onnx::Conv_579)
%onnx::Conv_585 = Identity(%onnx::Conv_579)
%onnx::Conv_582 = Identity(%onnx::Conv_579)
%onnx::Conv_576 = Identity(%onnx::Conv_552)
%onnx::Conv_573 = Identity(%onnx::Conv_552)
%onnx::Conv_570 = Identity(%onnx::Conv_552)
%onnx::Conv_567 = Identity(%onnx::Conv_552)
%onnx::Conv_564 = Identity(%onnx::Conv_552)
%onnx::Conv_561 = Identity(%onnx::Conv_552)
%onnx::Conv_558 = Identity(%onnx::Conv_552)
%onnx::Conv_555 = Identity(%onnx::Conv_552)
%onnx::Conv_549 = Identity(%onnx::Conv_534)
%onnx::Conv_546 = Identity(%onnx::Conv_534)
%onnx::Conv_543 = Identity(%onnx::Conv_534)
%onnx::Conv_540 = Identity(%onnx::Conv_534)
%onnx::Conv_537 = Identity(%onnx::Conv_534)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_533, %onnx::Conv_534)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_536, %onnx::Conv_537)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_539, %onnx::Conv_540)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_542, %onnx::Conv_543)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_545, %onnx::Conv_546)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_548, %onnx::Conv_549)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_551, %onnx::Conv_552)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_554, %onnx::Conv_555)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_557, %onnx::Conv_558)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_560, %onnx::Conv_561)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_563, %onnx::Conv_564)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_566, %onnx::Conv_567)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_569, %onnx::Conv_570)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_572, %onnx::Conv_573)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_575, %onnx::Conv_576)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_578, %onnx::Conv_579)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_581, %onnx::Conv_582)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_584, %onnx::Conv_585)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_587, %onnx::Conv_588)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_590, %onnx::Conv_591)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_593, %onnx::Conv_594)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_596, %onnx::Conv_597)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.8/Add_output_0, %onnx::Conv_599, %onnx::Conv_600)
%/cells.9/relu/Relu_output_0 = Relu(%/cells.9/conv/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/relu/Relu_output_0, %onnx::Conv_602, %onnx::Conv_603)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_605, %onnx::Conv_606)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_608, %onnx::Conv_609)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.9/relu/Relu_output_0)
%/cells.13/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_611, %onnx::Conv_612)
%/cells.13/relu/Relu_output_0 = Relu(%/cells.13/conv/Conv_output_0)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/relu/Relu_output_0, %onnx::Conv_614, %onnx::Conv_615)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_617, %onnx::Conv_618)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_620, %onnx::Conv_621)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/relu/Relu_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_623, %onnx::Conv_624)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%531 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %531
}
|
val_accuracy
| 0
| 62,723,200
| 2,238,724
|
{'zcp_synflow': 66.68541222793941, 'zcp_zen': 57.69160461425781, 'zcp_epe_nas': 21.59762150572456, 'zcp_fisher': 0.03316381946206093, 'zcp_flops': 62723200.0, 'zcp_grad_norm': 14.184907913208008, 'zcp_grasp': 0.03087902069091797, 'zcp_jacov': -16.05753035745849, 'zcp_l2_norm': 573.510498046875, 'zcp_nwot': 203.39606098535208, 'zcp_params': 2238724.0, 'zcp_plain': -0.010218581184744835, 'zcp_snip': 24.736854553222656, 'lat_1080ti_1': 0.2514357305656983, 'lat_1080ti_32': 0.18244258623463372, 'lat_1080ti_64': 0.09611816188791573, 'lat_2080ti_1': 0.26432965922513707, 'lat_2080ti_32': 0.17577826011878636, 'lat_2080ti_64': 0.0861150954267899, 'lat_essential_ph_1': 0.24528301886792453, 'lat_eyeriss': 0.269213136999615, 'lat_fpga': 0.4411482401369496, 'lat_gold_6226': 0.4073433056253166, 'lat_gold_6240': 0.4188259466034248, 'lat_pixel2': 0.21739130434782608, 'lat_pixel3': 0.28736805617583483, 'lat_raspi4': 0.3745718264291172, 'lat_samsung_a50': 0.1368421052631579, 'lat_samsung_s7': 0.13385826771653545, 'lat_silver_4114': 0.4320257681516836, 'lat_silver_4210r': 0.41309692325684194, 'lat_titan_rtx_1': 0.23809302098629703, 'lat_titan_rtx_32': 0.1641665057715928, 'lat_titan_rtx_64': 0.08232874067642955, 'lat_titanx_1': 0.1255208461811196, 'lat_titanx_32': 0.10209290806821635, 'lat_titanx_64': 0.08469541568378436, 'lat_titanxp_1': 0.25019919887900083, 'lat_titanxp_32': 0.1261326964106624, 'lat_titanxp_64': 0.07999711756562189}
| |
FBNet_4800
|
FBNet
|
4800
|
4800
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_660[FLOAT, 16x3x3x3]
%onnx::Conv_661[FLOAT, 16]
%onnx::Conv_663[FLOAT, 96x16x1x1]
%onnx::Conv_664[FLOAT, 96]
%onnx::Conv_666[FLOAT, 96x1x3x3]
%onnx::Conv_669[FLOAT, 16x96x1x1]
%onnx::Conv_672[FLOAT, 16x16x1x1]
%onnx::Conv_675[FLOAT, 16x1x5x5]
%onnx::Conv_678[FLOAT, 24x16x1x1]
%onnx::Conv_679[FLOAT, 24]
%onnx::Conv_681[FLOAT, 72x24x1x1]
%onnx::Conv_682[FLOAT, 72]
%onnx::Conv_684[FLOAT, 72x1x3x3]
%onnx::Conv_687[FLOAT, 24x72x1x1]
%onnx::Conv_690[FLOAT, 72x24x1x1]
%onnx::Conv_693[FLOAT, 72x1x3x3]
%onnx::Conv_696[FLOAT, 24x72x1x1]
%onnx::Conv_699[FLOAT, 24x24x1x1]
%onnx::Conv_702[FLOAT, 24x1x3x3]
%onnx::Conv_705[FLOAT, 24x24x1x1]
%onnx::Conv_708[FLOAT, 144x24x1x1]
%onnx::Conv_709[FLOAT, 144]
%onnx::Conv_711[FLOAT, 144x1x3x3]
%onnx::Conv_714[FLOAT, 32x144x1x1]
%onnx::Conv_715[FLOAT, 32]
%onnx::Conv_717[FLOAT, 96x32x1x1]
%onnx::Conv_720[FLOAT, 96x1x3x3]
%onnx::Conv_723[FLOAT, 32x96x1x1]
%onnx::Conv_726[FLOAT, 32x16x1x1]
%onnx::Conv_729[FLOAT, 32x1x5x5]
%onnx::Conv_732[FLOAT, 32x16x1x1]
%onnx::Conv_735[FLOAT, 96x32x1x1]
%onnx::Conv_738[FLOAT, 96x1x5x5]
%onnx::Conv_741[FLOAT, 64x96x1x1]
%onnx::Conv_742[FLOAT, 64]
%onnx::Conv_744[FLOAT, 64x32x1x1]
%onnx::Conv_747[FLOAT, 64x1x3x3]
%onnx::Conv_750[FLOAT, 64x32x1x1]
%onnx::Conv_753[FLOAT, 384x64x1x1]
%onnx::Conv_754[FLOAT, 384]
%onnx::Conv_756[FLOAT, 384x1x5x5]
%onnx::Conv_759[FLOAT, 64x384x1x1]
%onnx::Conv_762[FLOAT, 112x64x1x1]
%onnx::Conv_763[FLOAT, 112]
%onnx::Conv_765[FLOAT, 112x56x1x1]
%onnx::Conv_768[FLOAT, 112x1x3x3]
%onnx::Conv_771[FLOAT, 112x56x1x1]
%onnx::Conv_774[FLOAT, 672x112x1x1]
%onnx::Conv_775[FLOAT, 672]
%onnx::Conv_777[FLOAT, 672x1x3x3]
%onnx::Conv_780[FLOAT, 112x672x1x1]
%onnx::Conv_783[FLOAT, 672x112x1x1]
%onnx::Conv_786[FLOAT, 672x1x3x3]
%onnx::Conv_789[FLOAT, 112x672x1x1]
%onnx::Conv_792[FLOAT, 112x56x1x1]
%onnx::Conv_795[FLOAT, 112x1x5x5]
%onnx::Conv_798[FLOAT, 184x56x1x1]
%onnx::Conv_799[FLOAT, 184]
%onnx::Conv_801[FLOAT, 1104x184x1x1]
%onnx::Conv_802[FLOAT, 1104]
%onnx::Conv_804[FLOAT, 1104x1x3x3]
%onnx::Conv_807[FLOAT, 184x1104x1x1]
%onnx::Conv_810[FLOAT, 184x92x1x1]
%onnx::Conv_813[FLOAT, 184x1x3x3]
%onnx::Conv_816[FLOAT, 184x92x1x1]
%onnx::Conv_819[FLOAT, 1104x184x1x1]
%onnx::Conv_822[FLOAT, 1104x1x5x5]
%onnx::Conv_825[FLOAT, 184x1104x1x1]
%onnx::Conv_828[FLOAT, 184x92x1x1]
%onnx::Conv_831[FLOAT, 184x1x5x5]
%onnx::Conv_834[FLOAT, 352x92x1x1]
%onnx::Conv_835[FLOAT, 352]
%onnx::Conv_837[FLOAT, 1504x352x1x1]
%onnx::Conv_838[FLOAT, 1504]
) {
%onnx::Conv_832 = Identity(%onnx::Conv_799)
%onnx::Conv_829 = Identity(%onnx::Conv_799)
%onnx::Conv_826 = Identity(%onnx::Conv_799)
%onnx::Conv_823 = Identity(%onnx::Conv_802)
%onnx::Conv_820 = Identity(%onnx::Conv_802)
%onnx::Conv_817 = Identity(%onnx::Conv_799)
%onnx::Conv_814 = Identity(%onnx::Conv_799)
%onnx::Conv_811 = Identity(%onnx::Conv_799)
%onnx::Conv_808 = Identity(%onnx::Conv_799)
%onnx::Conv_805 = Identity(%onnx::Conv_802)
%onnx::Conv_796 = Identity(%onnx::Conv_763)
%onnx::Conv_793 = Identity(%onnx::Conv_763)
%onnx::Conv_790 = Identity(%onnx::Conv_763)
%onnx::Conv_787 = Identity(%onnx::Conv_775)
%onnx::Conv_784 = Identity(%onnx::Conv_775)
%onnx::Conv_781 = Identity(%onnx::Conv_763)
%onnx::Conv_778 = Identity(%onnx::Conv_775)
%onnx::Conv_772 = Identity(%onnx::Conv_763)
%onnx::Conv_769 = Identity(%onnx::Conv_763)
%onnx::Conv_766 = Identity(%onnx::Conv_763)
%onnx::Conv_760 = Identity(%onnx::Conv_742)
%onnx::Conv_757 = Identity(%onnx::Conv_754)
%onnx::Conv_751 = Identity(%onnx::Conv_742)
%onnx::Conv_748 = Identity(%onnx::Conv_742)
%onnx::Conv_745 = Identity(%onnx::Conv_742)
%onnx::Conv_739 = Identity(%onnx::Conv_664)
%onnx::Conv_736 = Identity(%onnx::Conv_664)
%onnx::Conv_733 = Identity(%onnx::Conv_715)
%onnx::Conv_730 = Identity(%onnx::Conv_715)
%onnx::Conv_727 = Identity(%onnx::Conv_715)
%onnx::Conv_724 = Identity(%onnx::Conv_715)
%onnx::Conv_721 = Identity(%onnx::Conv_664)
%onnx::Conv_718 = Identity(%onnx::Conv_664)
%onnx::Conv_712 = Identity(%onnx::Conv_709)
%onnx::Conv_706 = Identity(%onnx::Conv_679)
%onnx::Conv_703 = Identity(%onnx::Conv_679)
%onnx::Conv_700 = Identity(%onnx::Conv_679)
%onnx::Conv_697 = Identity(%onnx::Conv_679)
%onnx::Conv_694 = Identity(%onnx::Conv_682)
%onnx::Conv_691 = Identity(%onnx::Conv_682)
%onnx::Conv_688 = Identity(%onnx::Conv_679)
%onnx::Conv_685 = Identity(%onnx::Conv_682)
%onnx::Conv_676 = Identity(%onnx::Conv_661)
%onnx::Conv_673 = Identity(%onnx::Conv_661)
%onnx::Conv_670 = Identity(%onnx::Conv_661)
%onnx::Conv_667 = Identity(%onnx::Conv_664)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_660, %onnx::Conv_661)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.13/relu/Relu_output_0 = Relu(%/cells.13/conv/Conv_output_0)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/relu/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/relu/Relu_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%658 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %658
}
|
val_accuracy
| 0
| 77,132,672
| 2,108,748
|
{'zcp_synflow': 71.50479394864678, 'zcp_zen': 64.81527709960938, 'zcp_epe_nas': 7.948915027465914, 'zcp_fisher': 0.148245170712471, 'zcp_flops': 77132672.0, 'zcp_grad_norm': 25.1986141204834, 'zcp_grasp': -0.1459980010986328, 'zcp_jacov': -16.069554004036185, 'zcp_l2_norm': 616.6945190429688, 'zcp_nwot': 215.0782349247878, 'zcp_params': 2108748.0, 'zcp_plain': 0.0007776473648846149, 'zcp_snip': 47.65724182128906, 'lat_1080ti_1': 0.5738038380607667, 'lat_1080ti_32': 0.4941703264228818, 'lat_1080ti_64': 0.4436942868110475, 'lat_2080ti_1': 0.5977849515161364, 'lat_2080ti_32': 0.5284964715128877, 'lat_2080ti_64': 0.49093864344058635, 'lat_essential_ph_1': 0.2641509433962264, 'lat_eyeriss': 0.502796471861075, 'lat_fpga': 0.6035385269990797, 'lat_gold_6226': 0.45299305323767003, 'lat_gold_6240': 0.5577289969536979, 'lat_pixel2': 0.41304347826086957, 'lat_pixel3': 0.45636903119045064, 'lat_raspi4': 0.498600909109322, 'lat_samsung_a50': 0.2631578947368421, 'lat_samsung_s7': 0.08661417322834646, 'lat_silver_4114': 0.5822853573192054, 'lat_silver_4210r': 0.6029244305020038, 'lat_titan_rtx_1': 0.5543076423233838, 'lat_titan_rtx_32': 0.4872335904613819, 'lat_titan_rtx_64': 0.474784627592657, 'lat_titanx_1': 0.2931014577517552, 'lat_titanx_32': 0.4673133526384347, 'lat_titanx_64': 0.45278418832578876, 'lat_titanxp_1': 0.5249681557969109, 'lat_titanxp_32': 0.4707552155760541, 'lat_titanxp_64': 0.4685105674772332}
| |
FBNet_3759
|
FBNet
|
3759
|
3759
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_560[FLOAT, 16x3x3x3]
%onnx::Conv_561[FLOAT, 16]
%onnx::Conv_563[FLOAT, 16x16x1x1]
%onnx::Conv_566[FLOAT, 16x1x3x3]
%onnx::Conv_569[FLOAT, 16x16x1x1]
%onnx::Conv_572[FLOAT, 96x16x1x1]
%onnx::Conv_573[FLOAT, 96]
%onnx::Conv_575[FLOAT, 96x1x5x5]
%onnx::Conv_578[FLOAT, 24x96x1x1]
%onnx::Conv_579[FLOAT, 24]
%onnx::Conv_581[FLOAT, 72x24x1x1]
%onnx::Conv_582[FLOAT, 72]
%onnx::Conv_584[FLOAT, 72x1x3x3]
%onnx::Conv_587[FLOAT, 24x72x1x1]
%onnx::Conv_590[FLOAT, 144x24x1x1]
%onnx::Conv_591[FLOAT, 144]
%onnx::Conv_593[FLOAT, 144x1x5x5]
%onnx::Conv_596[FLOAT, 24x144x1x1]
%onnx::Conv_599[FLOAT, 24x24x1x1]
%onnx::Conv_602[FLOAT, 24x1x5x5]
%onnx::Conv_605[FLOAT, 24x24x1x1]
%onnx::Conv_608[FLOAT, 24x24x1x1]
%onnx::Conv_611[FLOAT, 24x1x5x5]
%onnx::Conv_614[FLOAT, 32x24x1x1]
%onnx::Conv_615[FLOAT, 32]
%onnx::Conv_617[FLOAT, 32x32x1x1]
%onnx::Conv_620[FLOAT, 32x1x5x5]
%onnx::Conv_623[FLOAT, 32x32x1x1]
%onnx::Conv_626[FLOAT, 32x16x1x1]
%onnx::Conv_629[FLOAT, 32x1x5x5]
%onnx::Conv_632[FLOAT, 32x16x1x1]
%onnx::Conv_635[FLOAT, 64x32x1x1]
%onnx::Conv_636[FLOAT, 64]
%onnx::Conv_638[FLOAT, 64x32x1x1]
%onnx::Conv_641[FLOAT, 64x1x3x3]
%onnx::Conv_644[FLOAT, 64x32x1x1]
%onnx::Conv_647[FLOAT, 384x64x1x1]
%onnx::Conv_648[FLOAT, 384]
%onnx::Conv_650[FLOAT, 384x1x3x3]
%onnx::Conv_653[FLOAT, 64x384x1x1]
%onnx::Conv_656[FLOAT, 64x32x1x1]
%onnx::Conv_659[FLOAT, 64x1x3x3]
%onnx::Conv_662[FLOAT, 64x32x1x1]
%onnx::Conv_665[FLOAT, 384x64x1x1]
%onnx::Conv_668[FLOAT, 384x1x5x5]
%onnx::Conv_671[FLOAT, 112x384x1x1]
%onnx::Conv_672[FLOAT, 112]
%onnx::Conv_674[FLOAT, 672x112x1x1]
%onnx::Conv_675[FLOAT, 672]
%onnx::Conv_677[FLOAT, 672x1x5x5]
%onnx::Conv_680[FLOAT, 112x672x1x1]
%onnx::Conv_683[FLOAT, 112x112x1x1]
%onnx::Conv_686[FLOAT, 112x1x5x5]
%onnx::Conv_689[FLOAT, 112x112x1x1]
%onnx::Conv_692[FLOAT, 112x112x1x1]
%onnx::Conv_695[FLOAT, 112x1x3x3]
%onnx::Conv_698[FLOAT, 184x112x1x1]
%onnx::Conv_699[FLOAT, 184]
%onnx::Conv_701[FLOAT, 184x184x1x1]
%onnx::Conv_704[FLOAT, 184x1x3x3]
%onnx::Conv_707[FLOAT, 184x184x1x1]
%onnx::Conv_710[FLOAT, 184x184x1x1]
%onnx::Conv_713[FLOAT, 184x1x3x3]
%onnx::Conv_716[FLOAT, 184x184x1x1]
%onnx::Conv_719[FLOAT, 352x184x1x1]
%onnx::Conv_720[FLOAT, 352]
%onnx::Conv_722[FLOAT, 1504x352x1x1]
%onnx::Conv_723[FLOAT, 1504]
) {
%onnx::Conv_717 = Identity(%onnx::Conv_699)
%onnx::Conv_714 = Identity(%onnx::Conv_699)
%onnx::Conv_711 = Identity(%onnx::Conv_699)
%onnx::Conv_708 = Identity(%onnx::Conv_699)
%onnx::Conv_705 = Identity(%onnx::Conv_699)
%onnx::Conv_702 = Identity(%onnx::Conv_699)
%onnx::Conv_696 = Identity(%onnx::Conv_672)
%onnx::Conv_693 = Identity(%onnx::Conv_672)
%onnx::Conv_690 = Identity(%onnx::Conv_672)
%onnx::Conv_687 = Identity(%onnx::Conv_672)
%onnx::Conv_684 = Identity(%onnx::Conv_672)
%onnx::Conv_681 = Identity(%onnx::Conv_672)
%onnx::Conv_678 = Identity(%onnx::Conv_675)
%onnx::Conv_669 = Identity(%onnx::Conv_648)
%onnx::Conv_666 = Identity(%onnx::Conv_648)
%onnx::Conv_663 = Identity(%onnx::Conv_636)
%onnx::Conv_660 = Identity(%onnx::Conv_636)
%onnx::Conv_657 = Identity(%onnx::Conv_636)
%onnx::Conv_654 = Identity(%onnx::Conv_636)
%onnx::Conv_651 = Identity(%onnx::Conv_648)
%onnx::Conv_645 = Identity(%onnx::Conv_636)
%onnx::Conv_642 = Identity(%onnx::Conv_636)
%onnx::Conv_639 = Identity(%onnx::Conv_636)
%onnx::Conv_633 = Identity(%onnx::Conv_615)
%onnx::Conv_630 = Identity(%onnx::Conv_615)
%onnx::Conv_627 = Identity(%onnx::Conv_615)
%onnx::Conv_624 = Identity(%onnx::Conv_615)
%onnx::Conv_621 = Identity(%onnx::Conv_615)
%onnx::Conv_618 = Identity(%onnx::Conv_615)
%onnx::Conv_612 = Identity(%onnx::Conv_579)
%onnx::Conv_609 = Identity(%onnx::Conv_579)
%onnx::Conv_606 = Identity(%onnx::Conv_579)
%onnx::Conv_603 = Identity(%onnx::Conv_579)
%onnx::Conv_600 = Identity(%onnx::Conv_579)
%onnx::Conv_597 = Identity(%onnx::Conv_579)
%onnx::Conv_594 = Identity(%onnx::Conv_591)
%onnx::Conv_588 = Identity(%onnx::Conv_579)
%onnx::Conv_585 = Identity(%onnx::Conv_582)
%onnx::Conv_576 = Identity(%onnx::Conv_573)
%onnx::Conv_570 = Identity(%onnx::Conv_561)
%onnx::Conv_567 = Identity(%onnx::Conv_561)
%onnx::Conv_564 = Identity(%onnx::Conv_561)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_560, %onnx::Conv_561)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_563, %onnx::Conv_564)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_566, %onnx::Conv_567)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_569, %onnx::Conv_570)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_572, %onnx::Conv_573)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_575, %onnx::Conv_576)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_578, %onnx::Conv_579)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_581, %onnx::Conv_582)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_584, %onnx::Conv_585)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_587, %onnx::Conv_588)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_590, %onnx::Conv_591)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_593, %onnx::Conv_594)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_596, %onnx::Conv_597)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_599, %onnx::Conv_600)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_602, %onnx::Conv_603)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_605, %onnx::Conv_606)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_608, %onnx::Conv_609)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_611, %onnx::Conv_612)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_614, %onnx::Conv_615)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_617, %onnx::Conv_618)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_620, %onnx::Conv_621)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_623, %onnx::Conv_624)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.8/Add_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.9/relu/Relu_output_0 = Relu(%/cells.9/conv/Conv_output_0)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/relu/Relu_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/relu/Relu_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.21/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.21/relu/Relu_output_0 = Relu(%/cells.21/conv/Conv_output_0)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/relu/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%558 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %558
}
|
val_accuracy
| 0
| 64,032,896
| 1,301,132
|
{'zcp_synflow': 72.33772937209817, 'zcp_zen': 58.111473083496094, 'zcp_epe_nas': 7.00796435638606, 'zcp_fisher': 0.05185583233833313, 'zcp_flops': 64032896.0, 'zcp_grad_norm': 18.16594123840332, 'zcp_grasp': 0.004061698913574219, 'zcp_jacov': -16.056613274196366, 'zcp_l2_norm': 506.10198974609375, 'zcp_nwot': 212.82835716831204, 'zcp_params': 1301132.0, 'zcp_plain': -0.004001906607300043, 'zcp_snip': 30.028751373291016, 'lat_1080ti_1': 0.3621845692339121, 'lat_1080ti_32': 0.4338997930025798, 'lat_1080ti_64': 0.4741536102508738, 'lat_2080ti_1': 0.38342160399723024, 'lat_2080ti_32': 0.45700391370775606, 'lat_2080ti_64': 0.47013062016261653, 'lat_essential_ph_1': 0.20754716981132076, 'lat_eyeriss': 0.38574807914473813, 'lat_fpga': 0.36186781978844046, 'lat_gold_6226': 0.1763586233489141, 'lat_gold_6240': 0.2587710746764246, 'lat_pixel2': 0.17391304347826086, 'lat_pixel3': 0.4410523038557249, 'lat_raspi4': 0.4069783148922776, 'lat_samsung_a50': 0.12631578947368421, 'lat_samsung_s7': 0.09448818897637795, 'lat_silver_4114': 0.26513020494433936, 'lat_silver_4210r': 0.29742920643933796, 'lat_titan_rtx_1': 0.3347985824686791, 'lat_titan_rtx_32': 0.4196622285275963, 'lat_titan_rtx_64': 0.45491195115261, 'lat_titanx_1': 0.18672207618832784, 'lat_titanx_32': 0.431916951756703, 'lat_titanx_64': 0.4761785093386063, 'lat_titanxp_1': 0.31066618893063846, 'lat_titanxp_32': 0.4290942604479259, 'lat_titanxp_64': 0.48104033381225636}
| |
FBNet_149
|
FBNet
|
149
|
149
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_615[FLOAT, 16x3x3x3]
%onnx::Conv_616[FLOAT, 16]
%onnx::Conv_618[FLOAT, 48x16x1x1]
%onnx::Conv_619[FLOAT, 48]
%onnx::Conv_621[FLOAT, 48x1x3x3]
%onnx::Conv_624[FLOAT, 16x48x1x1]
%onnx::Conv_627[FLOAT, 16x8x1x1]
%onnx::Conv_630[FLOAT, 16x1x5x5]
%onnx::Conv_633[FLOAT, 24x8x1x1]
%onnx::Conv_634[FLOAT, 24]
%onnx::Conv_636[FLOAT, 24x24x1x1]
%onnx::Conv_639[FLOAT, 24x1x5x5]
%onnx::Conv_642[FLOAT, 24x24x1x1]
%onnx::Conv_645[FLOAT, 24x12x1x1]
%onnx::Conv_648[FLOAT, 24x1x3x3]
%onnx::Conv_651[FLOAT, 24x12x1x1]
%onnx::Conv_654[FLOAT, 32x24x1x1]
%onnx::Conv_655[FLOAT, 32]
%onnx::Conv_657[FLOAT, 96x32x1x1]
%onnx::Conv_658[FLOAT, 96]
%onnx::Conv_660[FLOAT, 96x1x3x3]
%onnx::Conv_663[FLOAT, 32x96x1x1]
%onnx::Conv_666[FLOAT, 32x16x1x1]
%onnx::Conv_669[FLOAT, 32x1x3x3]
%onnx::Conv_672[FLOAT, 32x16x1x1]
%onnx::Conv_675[FLOAT, 96x32x1x1]
%onnx::Conv_678[FLOAT, 96x1x5x5]
%onnx::Conv_681[FLOAT, 32x96x1x1]
%onnx::Conv_684[FLOAT, 32x16x1x1]
%onnx::Conv_687[FLOAT, 32x1x3x3]
%onnx::Conv_690[FLOAT, 64x16x1x1]
%onnx::Conv_691[FLOAT, 64]
%onnx::Conv_693[FLOAT, 192x64x1x1]
%onnx::Conv_694[FLOAT, 192]
%onnx::Conv_696[FLOAT, 192x1x3x3]
%onnx::Conv_699[FLOAT, 64x192x1x1]
%onnx::Conv_702[FLOAT, 64x64x1x1]
%onnx::Conv_705[FLOAT, 64x1x5x5]
%onnx::Conv_708[FLOAT, 64x64x1x1]
%onnx::Conv_711[FLOAT, 192x64x1x1]
%onnx::Conv_714[FLOAT, 192x1x3x3]
%onnx::Conv_717[FLOAT, 112x192x1x1]
%onnx::Conv_718[FLOAT, 112]
%onnx::Conv_720[FLOAT, 672x112x1x1]
%onnx::Conv_721[FLOAT, 672]
%onnx::Conv_723[FLOAT, 672x1x5x5]
%onnx::Conv_726[FLOAT, 112x672x1x1]
%onnx::Conv_729[FLOAT, 112x112x1x1]
%onnx::Conv_732[FLOAT, 112x1x3x3]
%onnx::Conv_735[FLOAT, 112x112x1x1]
%onnx::Conv_738[FLOAT, 112x112x1x1]
%onnx::Conv_741[FLOAT, 112x1x3x3]
%onnx::Conv_744[FLOAT, 184x112x1x1]
%onnx::Conv_745[FLOAT, 184]
%onnx::Conv_747[FLOAT, 552x184x1x1]
%onnx::Conv_748[FLOAT, 552]
%onnx::Conv_750[FLOAT, 552x1x3x3]
%onnx::Conv_753[FLOAT, 184x552x1x1]
%onnx::Conv_756[FLOAT, 184x92x1x1]
%onnx::Conv_759[FLOAT, 184x1x5x5]
%onnx::Conv_762[FLOAT, 184x92x1x1]
%onnx::Conv_765[FLOAT, 552x184x1x1]
%onnx::Conv_768[FLOAT, 552x1x3x3]
%onnx::Conv_771[FLOAT, 184x552x1x1]
%onnx::Conv_774[FLOAT, 184x184x1x1]
%onnx::Conv_777[FLOAT, 184x1x3x3]
%onnx::Conv_780[FLOAT, 352x184x1x1]
%onnx::Conv_781[FLOAT, 352]
%onnx::Conv_783[FLOAT, 1504x352x1x1]
%onnx::Conv_784[FLOAT, 1504]
) {
%onnx::Conv_778 = Identity(%onnx::Conv_745)
%onnx::Conv_775 = Identity(%onnx::Conv_745)
%onnx::Conv_772 = Identity(%onnx::Conv_745)
%onnx::Conv_769 = Identity(%onnx::Conv_748)
%onnx::Conv_766 = Identity(%onnx::Conv_748)
%onnx::Conv_763 = Identity(%onnx::Conv_745)
%onnx::Conv_760 = Identity(%onnx::Conv_745)
%onnx::Conv_757 = Identity(%onnx::Conv_745)
%onnx::Conv_754 = Identity(%onnx::Conv_745)
%onnx::Conv_751 = Identity(%onnx::Conv_748)
%onnx::Conv_742 = Identity(%onnx::Conv_718)
%onnx::Conv_739 = Identity(%onnx::Conv_718)
%onnx::Conv_736 = Identity(%onnx::Conv_718)
%onnx::Conv_733 = Identity(%onnx::Conv_718)
%onnx::Conv_730 = Identity(%onnx::Conv_718)
%onnx::Conv_727 = Identity(%onnx::Conv_718)
%onnx::Conv_724 = Identity(%onnx::Conv_721)
%onnx::Conv_715 = Identity(%onnx::Conv_694)
%onnx::Conv_712 = Identity(%onnx::Conv_694)
%onnx::Conv_709 = Identity(%onnx::Conv_691)
%onnx::Conv_706 = Identity(%onnx::Conv_691)
%onnx::Conv_703 = Identity(%onnx::Conv_691)
%onnx::Conv_700 = Identity(%onnx::Conv_691)
%onnx::Conv_697 = Identity(%onnx::Conv_694)
%onnx::Conv_688 = Identity(%onnx::Conv_655)
%onnx::Conv_685 = Identity(%onnx::Conv_655)
%onnx::Conv_682 = Identity(%onnx::Conv_655)
%onnx::Conv_679 = Identity(%onnx::Conv_658)
%onnx::Conv_676 = Identity(%onnx::Conv_658)
%onnx::Conv_673 = Identity(%onnx::Conv_655)
%onnx::Conv_670 = Identity(%onnx::Conv_655)
%onnx::Conv_667 = Identity(%onnx::Conv_655)
%onnx::Conv_664 = Identity(%onnx::Conv_655)
%onnx::Conv_661 = Identity(%onnx::Conv_658)
%onnx::Conv_652 = Identity(%onnx::Conv_634)
%onnx::Conv_649 = Identity(%onnx::Conv_634)
%onnx::Conv_646 = Identity(%onnx::Conv_634)
%onnx::Conv_643 = Identity(%onnx::Conv_634)
%onnx::Conv_640 = Identity(%onnx::Conv_634)
%onnx::Conv_637 = Identity(%onnx::Conv_634)
%onnx::Conv_631 = Identity(%onnx::Conv_616)
%onnx::Conv_628 = Identity(%onnx::Conv_616)
%onnx::Conv_625 = Identity(%onnx::Conv_616)
%onnx::Conv_622 = Identity(%onnx::Conv_619)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_615, %onnx::Conv_616)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_618, %onnx::Conv_619)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_621, %onnx::Conv_622)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_624, %onnx::Conv_625)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_627, %onnx::Conv_628)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_630, %onnx::Conv_631)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.5/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.3/Add_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.5/relu/Relu_output_0 = Relu(%/cells.5/conv/Conv_output_0)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/relu/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/relu/Relu_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%613 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %613
}
|
val_accuracy
| 0
| 48,056,192
| 1,578,612
|
{'zcp_synflow': 69.21700701471585, 'zcp_zen': 59.47916030883789, 'zcp_epe_nas': 20.43496421938334, 'zcp_fisher': 0.07152798771858215, 'zcp_flops': 48056192.0, 'zcp_grad_norm': 19.94756507873535, 'zcp_grasp': 0.020165443420410156, 'zcp_jacov': -16.074270609134032, 'zcp_l2_norm': 540.6394653320312, 'zcp_nwot': 203.71622746915907, 'zcp_params': 1578612.0, 'zcp_plain': -0.005751880817115307, 'zcp_snip': 31.680438995361328, 'lat_1080ti_1': 0.3723660715864311, 'lat_1080ti_32': 0.3060280127730886, 'lat_1080ti_64': 0.19813826897134365, 'lat_2080ti_1': 0.42632295382770374, 'lat_2080ti_32': 0.30243656144029124, 'lat_2080ti_64': 0.18051891564015776, 'lat_essential_ph_1': 0.2830188679245283, 'lat_eyeriss': 0.17797480424696968, 'lat_fpga': 0.23578305742190192, 'lat_gold_6226': 0.19705569886447932, 'lat_gold_6240': 0.3111858849728004, 'lat_pixel2': 0.13043478260869565, 'lat_pixel3': 0.19362951371811135, 'lat_raspi4': 0.19922676472974038, 'lat_samsung_a50': 0.08421052631578947, 'lat_samsung_s7': 0.12598425196850394, 'lat_silver_4114': 0.33769334643442417, 'lat_silver_4210r': 0.3278062501362819, 'lat_titan_rtx_1': 0.41677472483753153, 'lat_titan_rtx_32': 0.3238051763916883, 'lat_titan_rtx_64': 0.2048049593700094, 'lat_titanx_1': 0.21978154008738499, 'lat_titanx_32': 0.23205260487348506, 'lat_titanx_64': 0.1697090918610001, 'lat_titanxp_1': 0.42157455321777715, 'lat_titanxp_32': 0.2818795868006173, 'lat_titanxp_64': 0.1963734150707593}
| |
FBNet_516
|
FBNet
|
516
|
516
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_614[FLOAT, 16x3x3x3]
%onnx::Conv_615[FLOAT, 16]
%onnx::Conv_617[FLOAT, 16x16x1x1]
%onnx::Conv_620[FLOAT, 16x1x3x3]
%onnx::Conv_623[FLOAT, 16x16x1x1]
%onnx::Conv_626[FLOAT, 16x16x1x1]
%onnx::Conv_629[FLOAT, 16x1x5x5]
%onnx::Conv_632[FLOAT, 24x16x1x1]
%onnx::Conv_633[FLOAT, 24]
%onnx::Conv_635[FLOAT, 24x12x1x1]
%onnx::Conv_638[FLOAT, 24x1x3x3]
%onnx::Conv_641[FLOAT, 24x12x1x1]
%onnx::Conv_644[FLOAT, 24x12x1x1]
%onnx::Conv_647[FLOAT, 24x1x3x3]
%onnx::Conv_650[FLOAT, 24x12x1x1]
%onnx::Conv_653[FLOAT, 72x24x1x1]
%onnx::Conv_654[FLOAT, 72]
%onnx::Conv_656[FLOAT, 72x1x5x5]
%onnx::Conv_659[FLOAT, 24x72x1x1]
%onnx::Conv_662[FLOAT, 72x24x1x1]
%onnx::Conv_665[FLOAT, 72x1x5x5]
%onnx::Conv_668[FLOAT, 32x72x1x1]
%onnx::Conv_669[FLOAT, 32]
%onnx::Conv_671[FLOAT, 192x32x1x1]
%onnx::Conv_672[FLOAT, 192]
%onnx::Conv_674[FLOAT, 192x1x3x3]
%onnx::Conv_677[FLOAT, 32x192x1x1]
%onnx::Conv_680[FLOAT, 192x32x1x1]
%onnx::Conv_683[FLOAT, 192x1x3x3]
%onnx::Conv_686[FLOAT, 32x192x1x1]
%onnx::Conv_689[FLOAT, 96x32x1x1]
%onnx::Conv_690[FLOAT, 96]
%onnx::Conv_692[FLOAT, 96x1x5x5]
%onnx::Conv_695[FLOAT, 64x96x1x1]
%onnx::Conv_696[FLOAT, 64]
%onnx::Conv_698[FLOAT, 64x64x1x1]
%onnx::Conv_701[FLOAT, 64x1x3x3]
%onnx::Conv_704[FLOAT, 64x64x1x1]
%onnx::Conv_707[FLOAT, 384x64x1x1]
%onnx::Conv_708[FLOAT, 384]
%onnx::Conv_710[FLOAT, 384x1x3x3]
%onnx::Conv_713[FLOAT, 64x384x1x1]
%onnx::Conv_716[FLOAT, 64x32x1x1]
%onnx::Conv_719[FLOAT, 64x1x5x5]
%onnx::Conv_722[FLOAT, 64x32x1x1]
%onnx::Conv_725[FLOAT, 64x64x1x1]
%onnx::Conv_728[FLOAT, 64x1x3x3]
%onnx::Conv_731[FLOAT, 112x64x1x1]
%onnx::Conv_732[FLOAT, 112]
%onnx::Conv_734[FLOAT, 672x112x1x1]
%onnx::Conv_735[FLOAT, 672]
%onnx::Conv_737[FLOAT, 672x1x5x5]
%onnx::Conv_740[FLOAT, 112x672x1x1]
%onnx::Conv_743[FLOAT, 112x112x1x1]
%onnx::Conv_746[FLOAT, 112x1x5x5]
%onnx::Conv_749[FLOAT, 112x112x1x1]
%onnx::Conv_752[FLOAT, 336x112x1x1]
%onnx::Conv_753[FLOAT, 336]
%onnx::Conv_755[FLOAT, 336x1x5x5]
%onnx::Conv_758[FLOAT, 112x336x1x1]
%onnx::Conv_761[FLOAT, 184x112x1x1]
%onnx::Conv_762[FLOAT, 184]
%onnx::Conv_764[FLOAT, 184x92x1x1]
%onnx::Conv_767[FLOAT, 184x1x3x3]
%onnx::Conv_770[FLOAT, 184x92x1x1]
%onnx::Conv_773[FLOAT, 184x92x1x1]
%onnx::Conv_776[FLOAT, 184x1x5x5]
%onnx::Conv_779[FLOAT, 352x92x1x1]
%onnx::Conv_780[FLOAT, 352]
%onnx::Conv_782[FLOAT, 1504x352x1x1]
%onnx::Conv_783[FLOAT, 1504]
) {
%onnx::Conv_777 = Identity(%onnx::Conv_762)
%onnx::Conv_774 = Identity(%onnx::Conv_762)
%onnx::Conv_771 = Identity(%onnx::Conv_762)
%onnx::Conv_768 = Identity(%onnx::Conv_762)
%onnx::Conv_765 = Identity(%onnx::Conv_762)
%onnx::Conv_759 = Identity(%onnx::Conv_732)
%onnx::Conv_756 = Identity(%onnx::Conv_753)
%onnx::Conv_750 = Identity(%onnx::Conv_732)
%onnx::Conv_747 = Identity(%onnx::Conv_732)
%onnx::Conv_744 = Identity(%onnx::Conv_732)
%onnx::Conv_741 = Identity(%onnx::Conv_732)
%onnx::Conv_738 = Identity(%onnx::Conv_735)
%onnx::Conv_729 = Identity(%onnx::Conv_696)
%onnx::Conv_726 = Identity(%onnx::Conv_696)
%onnx::Conv_723 = Identity(%onnx::Conv_696)
%onnx::Conv_720 = Identity(%onnx::Conv_696)
%onnx::Conv_717 = Identity(%onnx::Conv_696)
%onnx::Conv_714 = Identity(%onnx::Conv_696)
%onnx::Conv_711 = Identity(%onnx::Conv_708)
%onnx::Conv_705 = Identity(%onnx::Conv_696)
%onnx::Conv_702 = Identity(%onnx::Conv_696)
%onnx::Conv_699 = Identity(%onnx::Conv_696)
%onnx::Conv_693 = Identity(%onnx::Conv_690)
%onnx::Conv_687 = Identity(%onnx::Conv_669)
%onnx::Conv_684 = Identity(%onnx::Conv_672)
%onnx::Conv_681 = Identity(%onnx::Conv_672)
%onnx::Conv_678 = Identity(%onnx::Conv_669)
%onnx::Conv_675 = Identity(%onnx::Conv_672)
%onnx::Conv_666 = Identity(%onnx::Conv_654)
%onnx::Conv_663 = Identity(%onnx::Conv_654)
%onnx::Conv_660 = Identity(%onnx::Conv_633)
%onnx::Conv_657 = Identity(%onnx::Conv_654)
%onnx::Conv_651 = Identity(%onnx::Conv_633)
%onnx::Conv_648 = Identity(%onnx::Conv_633)
%onnx::Conv_645 = Identity(%onnx::Conv_633)
%onnx::Conv_642 = Identity(%onnx::Conv_633)
%onnx::Conv_639 = Identity(%onnx::Conv_633)
%onnx::Conv_636 = Identity(%onnx::Conv_633)
%onnx::Conv_630 = Identity(%onnx::Conv_615)
%onnx::Conv_627 = Identity(%onnx::Conv_615)
%onnx::Conv_624 = Identity(%onnx::Conv_615)
%onnx::Conv_621 = Identity(%onnx::Conv_615)
%onnx::Conv_618 = Identity(%onnx::Conv_615)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_614, %onnx::Conv_615)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_617, %onnx::Conv_618)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_620, %onnx::Conv_621)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_623, %onnx::Conv_624)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_626, %onnx::Conv_627)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_629, %onnx::Conv_630)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_632, %onnx::Conv_633)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.16/Add_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.17/relu/Relu_output_0 = Relu(%/cells.17/conv/Conv_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/relu/Relu_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.17/relu/Relu_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%612 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %612
}
|
val_accuracy
| 0
| 56,445,184
| 1,219,556
|
{'zcp_synflow': 70.13072677096852, 'zcp_zen': 59.05784606933594, 'zcp_epe_nas': 39.48388292964794, 'zcp_fisher': 0.056797437369823456, 'zcp_flops': 56445184.0, 'zcp_grad_norm': 18.761117935180664, 'zcp_grasp': 0.007962226867675781, 'zcp_jacov': -16.05599099703403, 'zcp_l2_norm': 503.7805480957031, 'zcp_nwot': 210.02637077963408, 'zcp_params': 1219556.0, 'zcp_plain': -0.0047215670347213745, 'zcp_snip': 31.78639030456543, 'lat_1080ti_1': 0.39130308450371587, 'lat_1080ti_32': 0.3712192598007674, 'lat_1080ti_64': 0.2910872878819042, 'lat_2080ti_1': 0.4312277482485872, 'lat_2080ti_32': 0.3694533465666882, 'lat_2080ti_64': 0.29688394107361965, 'lat_essential_ph_1': 0.2830188679245283, 'lat_eyeriss': 0.2748885995635672, 'lat_fpga': 0.298884809863545, 'lat_gold_6226': 0.19213884270226822, 'lat_gold_6240': 0.2533067988562345, 'lat_pixel2': 0.15217391304347827, 'lat_pixel3': 0.312183808579847, 'lat_raspi4': 0.2738586218401447, 'lat_samsung_a50': 0.09473684210526316, 'lat_samsung_s7': 0.13385826771653545, 'lat_silver_4114': 0.32677485346917984, 'lat_silver_4210r': 0.3367823465672381, 'lat_titan_rtx_1': 0.42171953382588045, 'lat_titan_rtx_32': 0.3771996700196304, 'lat_titan_rtx_64': 0.2934355856690883, 'lat_titanx_1': 0.22293237437697086, 'lat_titanx_32': 0.2943759542202068, 'lat_titanx_64': 0.28013214983144863, 'lat_titanxp_1': 0.4012719651987773, 'lat_titanxp_32': 0.3218536285799875, 'lat_titanxp_64': 0.27775212287984036}
| |
FBNet_4018
|
FBNet
|
4018
|
4018
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_622[FLOAT, 16x3x3x3]
%onnx::Conv_623[FLOAT, 16]
%onnx::Conv_625[FLOAT, 16x16x1x1]
%onnx::Conv_628[FLOAT, 16x1x5x5]
%onnx::Conv_631[FLOAT, 16x16x1x1]
%onnx::Conv_634[FLOAT, 16x8x1x1]
%onnx::Conv_637[FLOAT, 16x1x3x3]
%onnx::Conv_640[FLOAT, 24x8x1x1]
%onnx::Conv_641[FLOAT, 24]
%onnx::Conv_643[FLOAT, 24x24x1x1]
%onnx::Conv_646[FLOAT, 24x1x5x5]
%onnx::Conv_649[FLOAT, 24x24x1x1]
%onnx::Conv_652[FLOAT, 72x24x1x1]
%onnx::Conv_653[FLOAT, 72]
%onnx::Conv_655[FLOAT, 72x1x5x5]
%onnx::Conv_658[FLOAT, 24x72x1x1]
%onnx::Conv_661[FLOAT, 72x24x1x1]
%onnx::Conv_664[FLOAT, 72x1x3x3]
%onnx::Conv_667[FLOAT, 24x72x1x1]
%onnx::Conv_670[FLOAT, 144x24x1x1]
%onnx::Conv_671[FLOAT, 144]
%onnx::Conv_673[FLOAT, 144x1x3x3]
%onnx::Conv_676[FLOAT, 32x144x1x1]
%onnx::Conv_677[FLOAT, 32]
%onnx::Conv_679[FLOAT, 192x32x1x1]
%onnx::Conv_680[FLOAT, 192]
%onnx::Conv_682[FLOAT, 192x1x5x5]
%onnx::Conv_685[FLOAT, 32x192x1x1]
%onnx::Conv_688[FLOAT, 192x32x1x1]
%onnx::Conv_691[FLOAT, 192x1x5x5]
%onnx::Conv_694[FLOAT, 32x192x1x1]
%onnx::Conv_697[FLOAT, 32x32x1x1]
%onnx::Conv_700[FLOAT, 32x1x3x3]
%onnx::Conv_703[FLOAT, 64x32x1x1]
%onnx::Conv_704[FLOAT, 64]
%onnx::Conv_706[FLOAT, 192x64x1x1]
%onnx::Conv_709[FLOAT, 192x1x3x3]
%onnx::Conv_712[FLOAT, 64x192x1x1]
%onnx::Conv_715[FLOAT, 192x64x1x1]
%onnx::Conv_718[FLOAT, 192x1x5x5]
%onnx::Conv_721[FLOAT, 64x192x1x1]
%onnx::Conv_724[FLOAT, 192x64x1x1]
%onnx::Conv_727[FLOAT, 192x1x5x5]
%onnx::Conv_730[FLOAT, 64x192x1x1]
%onnx::Conv_733[FLOAT, 384x64x1x1]
%onnx::Conv_734[FLOAT, 384]
%onnx::Conv_736[FLOAT, 384x1x3x3]
%onnx::Conv_739[FLOAT, 112x384x1x1]
%onnx::Conv_740[FLOAT, 112]
%onnx::Conv_742[FLOAT, 112x56x1x1]
%onnx::Conv_745[FLOAT, 112x1x5x5]
%onnx::Conv_748[FLOAT, 112x56x1x1]
%onnx::Conv_751[FLOAT, 112x56x1x1]
%onnx::Conv_754[FLOAT, 112x1x5x5]
%onnx::Conv_757[FLOAT, 112x56x1x1]
%onnx::Conv_760[FLOAT, 112x56x1x1]
%onnx::Conv_763[FLOAT, 112x1x3x3]
%onnx::Conv_766[FLOAT, 112x56x1x1]
%onnx::Conv_769[FLOAT, 672x112x1x1]
%onnx::Conv_770[FLOAT, 672]
%onnx::Conv_772[FLOAT, 672x1x3x3]
%onnx::Conv_775[FLOAT, 184x672x1x1]
%onnx::Conv_776[FLOAT, 184]
%onnx::Conv_778[FLOAT, 552x184x1x1]
%onnx::Conv_779[FLOAT, 552]
%onnx::Conv_781[FLOAT, 552x1x5x5]
%onnx::Conv_784[FLOAT, 184x552x1x1]
%onnx::Conv_787[FLOAT, 552x184x1x1]
%onnx::Conv_790[FLOAT, 552x1x5x5]
%onnx::Conv_793[FLOAT, 184x552x1x1]
%onnx::Conv_796[FLOAT, 352x184x1x1]
%onnx::Conv_797[FLOAT, 352]
%onnx::Conv_799[FLOAT, 1504x352x1x1]
%onnx::Conv_800[FLOAT, 1504]
) {
%onnx::Conv_794 = Identity(%onnx::Conv_776)
%onnx::Conv_791 = Identity(%onnx::Conv_779)
%onnx::Conv_788 = Identity(%onnx::Conv_779)
%onnx::Conv_785 = Identity(%onnx::Conv_776)
%onnx::Conv_782 = Identity(%onnx::Conv_779)
%onnx::Conv_773 = Identity(%onnx::Conv_770)
%onnx::Conv_767 = Identity(%onnx::Conv_740)
%onnx::Conv_764 = Identity(%onnx::Conv_740)
%onnx::Conv_761 = Identity(%onnx::Conv_740)
%onnx::Conv_758 = Identity(%onnx::Conv_740)
%onnx::Conv_755 = Identity(%onnx::Conv_740)
%onnx::Conv_752 = Identity(%onnx::Conv_740)
%onnx::Conv_749 = Identity(%onnx::Conv_740)
%onnx::Conv_746 = Identity(%onnx::Conv_740)
%onnx::Conv_743 = Identity(%onnx::Conv_740)
%onnx::Conv_737 = Identity(%onnx::Conv_734)
%onnx::Conv_731 = Identity(%onnx::Conv_704)
%onnx::Conv_728 = Identity(%onnx::Conv_680)
%onnx::Conv_725 = Identity(%onnx::Conv_680)
%onnx::Conv_722 = Identity(%onnx::Conv_704)
%onnx::Conv_719 = Identity(%onnx::Conv_680)
%onnx::Conv_716 = Identity(%onnx::Conv_680)
%onnx::Conv_713 = Identity(%onnx::Conv_704)
%onnx::Conv_710 = Identity(%onnx::Conv_680)
%onnx::Conv_707 = Identity(%onnx::Conv_680)
%onnx::Conv_701 = Identity(%onnx::Conv_677)
%onnx::Conv_698 = Identity(%onnx::Conv_677)
%onnx::Conv_695 = Identity(%onnx::Conv_677)
%onnx::Conv_692 = Identity(%onnx::Conv_680)
%onnx::Conv_689 = Identity(%onnx::Conv_680)
%onnx::Conv_686 = Identity(%onnx::Conv_677)
%onnx::Conv_683 = Identity(%onnx::Conv_680)
%onnx::Conv_674 = Identity(%onnx::Conv_671)
%onnx::Conv_668 = Identity(%onnx::Conv_641)
%onnx::Conv_665 = Identity(%onnx::Conv_653)
%onnx::Conv_662 = Identity(%onnx::Conv_653)
%onnx::Conv_659 = Identity(%onnx::Conv_641)
%onnx::Conv_656 = Identity(%onnx::Conv_653)
%onnx::Conv_650 = Identity(%onnx::Conv_641)
%onnx::Conv_647 = Identity(%onnx::Conv_641)
%onnx::Conv_644 = Identity(%onnx::Conv_641)
%onnx::Conv_638 = Identity(%onnx::Conv_623)
%onnx::Conv_635 = Identity(%onnx::Conv_623)
%onnx::Conv_632 = Identity(%onnx::Conv_623)
%onnx::Conv_629 = Identity(%onnx::Conv_623)
%onnx::Conv_626 = Identity(%onnx::Conv_623)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_622, %onnx::Conv_623)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_625, %onnx::Conv_626)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_628, %onnx::Conv_629)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_631, %onnx::Conv_632)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_634, %onnx::Conv_635)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_637, %onnx::Conv_638)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_640, %onnx::Conv_641)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_643, %onnx::Conv_644)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_646, %onnx::Conv_647)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_649, %onnx::Conv_650)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_652, %onnx::Conv_653)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_655, %onnx::Conv_656)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_658, %onnx::Conv_659)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.21/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.21/relu/Relu_output_0 = Relu(%/cells.21/conv/Conv_output_0)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/relu/Relu_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%620 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %620
}
|
val_accuracy
| 0
| 66,867,328
| 1,665,660
|
{'zcp_synflow': 76.19661582437598, 'zcp_zen': 65.77266693115234, 'zcp_epe_nas': 18.663864327761324, 'zcp_fisher': 0.0853259414434433, 'zcp_flops': 66867328.0, 'zcp_grad_norm': 20.521465301513672, 'zcp_grasp': 0.022581100463867188, 'zcp_jacov': -16.05743803576454, 'zcp_l2_norm': 593.912353515625, 'zcp_nwot': 213.1684613355981, 'zcp_params': 1665660.0, 'zcp_plain': -0.004168059676885605, 'zcp_snip': 36.83818054199219, 'lat_1080ti_1': 0.5671673519390612, 'lat_1080ti_32': 0.436339112127943, 'lat_1080ti_64': 0.4307937648865651, 'lat_2080ti_1': 0.538128427232357, 'lat_2080ti_32': 0.46080562313530155, 'lat_2080ti_64': 0.4577957865254797, 'lat_essential_ph_1': 0.16981132075471697, 'lat_eyeriss': 0.42877707076449134, 'lat_fpga': 0.3573457089423009, 'lat_gold_6226': 0.33874691751449004, 'lat_gold_6240': 0.4021184932842864, 'lat_pixel2': 0.2391304347826087, 'lat_pixel3': 0.4293473898648011, 'lat_raspi4': 0.40352801939955535, 'lat_samsung_a50': 0.17894736842105263, 'lat_samsung_s7': 0.10236220472440945, 'lat_silver_4114': 0.41781029638155454, 'lat_silver_4210r': 0.4367866941063779, 'lat_titan_rtx_1': 0.5196088304079397, 'lat_titan_rtx_32': 0.4341768274753655, 'lat_titan_rtx_64': 0.4469424157919243, 'lat_titanx_1': 0.2673672799771127, 'lat_titanx_32': 0.4235423084783476, 'lat_titanx_64': 0.45074611677810267, 'lat_titanxp_1': 0.4828873257503776, 'lat_titanxp_32': 0.4259700442246364, 'lat_titanxp_64': 0.42424856524277493}
| |
FBNet_1601
|
FBNet
|
1601
|
1601
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_679[FLOAT, 16x3x3x3]
%onnx::Conv_680[FLOAT, 16]
%onnx::Conv_682[FLOAT, 16x16x1x1]
%onnx::Conv_685[FLOAT, 16x1x3x3]
%onnx::Conv_688[FLOAT, 16x16x1x1]
%onnx::Conv_691[FLOAT, 24x16x1x1]
%onnx::Conv_692[FLOAT, 24]
%onnx::Conv_694[FLOAT, 24x12x1x1]
%onnx::Conv_697[FLOAT, 24x1x3x3]
%onnx::Conv_700[FLOAT, 24x12x1x1]
%onnx::Conv_703[FLOAT, 72x24x1x1]
%onnx::Conv_704[FLOAT, 72]
%onnx::Conv_706[FLOAT, 72x1x5x5]
%onnx::Conv_709[FLOAT, 24x72x1x1]
%onnx::Conv_712[FLOAT, 24x12x1x1]
%onnx::Conv_715[FLOAT, 24x1x5x5]
%onnx::Conv_718[FLOAT, 24x12x1x1]
%onnx::Conv_721[FLOAT, 144x24x1x1]
%onnx::Conv_722[FLOAT, 144]
%onnx::Conv_724[FLOAT, 144x1x3x3]
%onnx::Conv_727[FLOAT, 32x144x1x1]
%onnx::Conv_728[FLOAT, 32]
%onnx::Conv_730[FLOAT, 96x32x1x1]
%onnx::Conv_731[FLOAT, 96]
%onnx::Conv_733[FLOAT, 96x1x3x3]
%onnx::Conv_736[FLOAT, 32x96x1x1]
%onnx::Conv_739[FLOAT, 192x32x1x1]
%onnx::Conv_740[FLOAT, 192]
%onnx::Conv_742[FLOAT, 192x1x5x5]
%onnx::Conv_745[FLOAT, 32x192x1x1]
%onnx::Conv_748[FLOAT, 32x32x1x1]
%onnx::Conv_751[FLOAT, 32x1x5x5]
%onnx::Conv_754[FLOAT, 32x32x1x1]
%onnx::Conv_757[FLOAT, 96x32x1x1]
%onnx::Conv_760[FLOAT, 96x1x3x3]
%onnx::Conv_763[FLOAT, 64x96x1x1]
%onnx::Conv_764[FLOAT, 64]
%onnx::Conv_766[FLOAT, 64x64x1x1]
%onnx::Conv_769[FLOAT, 64x1x3x3]
%onnx::Conv_772[FLOAT, 64x64x1x1]
%onnx::Conv_775[FLOAT, 64x64x1x1]
%onnx::Conv_778[FLOAT, 64x1x3x3]
%onnx::Conv_781[FLOAT, 64x64x1x1]
%onnx::Conv_784[FLOAT, 64x32x1x1]
%onnx::Conv_787[FLOAT, 64x1x3x3]
%onnx::Conv_790[FLOAT, 64x32x1x1]
%onnx::Conv_793[FLOAT, 112x64x1x1]
%onnx::Conv_794[FLOAT, 112]
%onnx::Conv_796[FLOAT, 672x112x1x1]
%onnx::Conv_797[FLOAT, 672]
%onnx::Conv_799[FLOAT, 672x1x3x3]
%onnx::Conv_802[FLOAT, 112x672x1x1]
%onnx::Conv_805[FLOAT, 112x112x1x1]
%onnx::Conv_808[FLOAT, 112x1x3x3]
%onnx::Conv_811[FLOAT, 112x112x1x1]
%onnx::Conv_814[FLOAT, 112x56x1x1]
%onnx::Conv_817[FLOAT, 112x1x3x3]
%onnx::Conv_820[FLOAT, 112x56x1x1]
%onnx::Conv_823[FLOAT, 112x112x1x1]
%onnx::Conv_826[FLOAT, 112x1x3x3]
%onnx::Conv_829[FLOAT, 184x112x1x1]
%onnx::Conv_830[FLOAT, 184]
%onnx::Conv_832[FLOAT, 552x184x1x1]
%onnx::Conv_833[FLOAT, 552]
%onnx::Conv_835[FLOAT, 552x1x3x3]
%onnx::Conv_838[FLOAT, 184x552x1x1]
%onnx::Conv_841[FLOAT, 184x184x1x1]
%onnx::Conv_844[FLOAT, 184x1x5x5]
%onnx::Conv_847[FLOAT, 184x184x1x1]
%onnx::Conv_850[FLOAT, 184x184x1x1]
%onnx::Conv_853[FLOAT, 184x1x3x3]
%onnx::Conv_856[FLOAT, 184x184x1x1]
%onnx::Conv_859[FLOAT, 184x92x1x1]
%onnx::Conv_862[FLOAT, 184x1x5x5]
%onnx::Conv_865[FLOAT, 352x92x1x1]
%onnx::Conv_866[FLOAT, 352]
%onnx::Conv_868[FLOAT, 1504x352x1x1]
%onnx::Conv_869[FLOAT, 1504]
) {
%onnx::Conv_863 = Identity(%onnx::Conv_830)
%onnx::Conv_860 = Identity(%onnx::Conv_830)
%onnx::Conv_857 = Identity(%onnx::Conv_830)
%onnx::Conv_854 = Identity(%onnx::Conv_830)
%onnx::Conv_851 = Identity(%onnx::Conv_830)
%onnx::Conv_848 = Identity(%onnx::Conv_830)
%onnx::Conv_845 = Identity(%onnx::Conv_830)
%onnx::Conv_842 = Identity(%onnx::Conv_830)
%onnx::Conv_839 = Identity(%onnx::Conv_830)
%onnx::Conv_836 = Identity(%onnx::Conv_833)
%onnx::Conv_827 = Identity(%onnx::Conv_794)
%onnx::Conv_824 = Identity(%onnx::Conv_794)
%onnx::Conv_821 = Identity(%onnx::Conv_794)
%onnx::Conv_818 = Identity(%onnx::Conv_794)
%onnx::Conv_815 = Identity(%onnx::Conv_794)
%onnx::Conv_812 = Identity(%onnx::Conv_794)
%onnx::Conv_809 = Identity(%onnx::Conv_794)
%onnx::Conv_806 = Identity(%onnx::Conv_794)
%onnx::Conv_803 = Identity(%onnx::Conv_794)
%onnx::Conv_800 = Identity(%onnx::Conv_797)
%onnx::Conv_791 = Identity(%onnx::Conv_764)
%onnx::Conv_788 = Identity(%onnx::Conv_764)
%onnx::Conv_785 = Identity(%onnx::Conv_764)
%onnx::Conv_782 = Identity(%onnx::Conv_764)
%onnx::Conv_779 = Identity(%onnx::Conv_764)
%onnx::Conv_776 = Identity(%onnx::Conv_764)
%onnx::Conv_773 = Identity(%onnx::Conv_764)
%onnx::Conv_770 = Identity(%onnx::Conv_764)
%onnx::Conv_767 = Identity(%onnx::Conv_764)
%onnx::Conv_761 = Identity(%onnx::Conv_731)
%onnx::Conv_758 = Identity(%onnx::Conv_731)
%onnx::Conv_755 = Identity(%onnx::Conv_728)
%onnx::Conv_752 = Identity(%onnx::Conv_728)
%onnx::Conv_749 = Identity(%onnx::Conv_728)
%onnx::Conv_746 = Identity(%onnx::Conv_728)
%onnx::Conv_743 = Identity(%onnx::Conv_740)
%onnx::Conv_737 = Identity(%onnx::Conv_728)
%onnx::Conv_734 = Identity(%onnx::Conv_731)
%onnx::Conv_725 = Identity(%onnx::Conv_722)
%onnx::Conv_719 = Identity(%onnx::Conv_692)
%onnx::Conv_716 = Identity(%onnx::Conv_692)
%onnx::Conv_713 = Identity(%onnx::Conv_692)
%onnx::Conv_710 = Identity(%onnx::Conv_692)
%onnx::Conv_707 = Identity(%onnx::Conv_704)
%onnx::Conv_701 = Identity(%onnx::Conv_692)
%onnx::Conv_698 = Identity(%onnx::Conv_692)
%onnx::Conv_695 = Identity(%onnx::Conv_692)
%onnx::Conv_689 = Identity(%onnx::Conv_680)
%onnx::Conv_686 = Identity(%onnx::Conv_680)
%onnx::Conv_683 = Identity(%onnx::Conv_680)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_679, %onnx::Conv_680)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.1/relu/Relu_output_0 = Relu(%/cells.1/conv/Conv_output_0)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/relu/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/relu/Relu_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.13/relu/Relu_output_0 = Relu(%/cells.13/conv/Conv_output_0)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/relu/Relu_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/relu/Relu_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_output_0 = Reshape(%/cells.16/nl/Relu_output_0, %/cells.16/shuffle/Constant_output_0)
%/cells.16/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.16/shuffle/Reshape_output_0)
%/cells.16/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.16/shuffle/Reshape_1_output_0 = Reshape(%/cells.16/shuffle/Transpose_output_0, %/cells.16/shuffle/Constant_1_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/shuffle/Reshape_1_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_823, %onnx::Conv_824)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_826, %onnx::Conv_827)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_829, %onnx::Conv_830)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_832, %onnx::Conv_833)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_835, %onnx::Conv_836)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_838, %onnx::Conv_839)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_841, %onnx::Conv_842)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_844, %onnx::Conv_845)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_847, %onnx::Conv_848)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_850, %onnx::Conv_851)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_853, %onnx::Conv_854)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_856, %onnx::Conv_857)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_859, %onnx::Conv_860)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_862, %onnx::Conv_863)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_865, %onnx::Conv_866)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_868, %onnx::Conv_869)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%677 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %677
}
|
val_accuracy
| 0
| 55,387,008
| 1,417,548
|
{'zcp_synflow': 78.13113515636103, 'zcp_zen': 66.14258575439453, 'zcp_epe_nas': 13.09920014071281, 'zcp_fisher': 0.15095022320747375, 'zcp_flops': 55387008.0, 'zcp_grad_norm': 25.411060333251953, 'zcp_grasp': -0.20779037475585938, 'zcp_jacov': -16.060828697769725, 'zcp_l2_norm': 584.6094360351562, 'zcp_nwot': 210.24583921526946, 'zcp_params': 1417548.0, 'zcp_plain': 0.0026208998169749975, 'zcp_snip': 39.523929595947266, 'lat_1080ti_1': 0.6204140783723668, 'lat_1080ti_32': 0.5272342861585899, 'lat_1080ti_64': 0.4000641834939449, 'lat_2080ti_1': 0.6828204511369197, 'lat_2080ti_32': 0.5601797272082886, 'lat_2080ti_64': 0.4480314337013254, 'lat_essential_ph_1': 0.16981132075471697, 'lat_eyeriss': 0.2800689490767059, 'lat_fpga': 0.2645979081526633, 'lat_gold_6226': 0.1644395077137422, 'lat_gold_6240': 0.369987728980098, 'lat_pixel2': 0.1956521739130435, 'lat_pixel3': 0.2869347498646531, 'lat_raspi4': 0.2731427859383723, 'lat_samsung_a50': 0.11578947368421053, 'lat_samsung_s7': 0.07874015748031496, 'lat_silver_4114': 0.5603425091221376, 'lat_silver_4210r': 0.4397975587315947, 'lat_titan_rtx_1': 0.6628832339896625, 'lat_titan_rtx_32': 0.554085383836952, 'lat_titan_rtx_64': 0.48030859003509896, 'lat_titanx_1': 0.3549737943401653, 'lat_titanx_32': 0.4935803510436706, 'lat_titanx_64': 0.3783828755812799, 'lat_titanxp_1': 0.6317818305112457, 'lat_titanxp_32': 0.526788947855743, 'lat_titanxp_64': 0.4327650001543177}
| |
FBNet_2878
|
FBNet
|
2878
|
2878
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_650[FLOAT, 16x3x3x3]
%onnx::Conv_651[FLOAT, 16]
%onnx::Conv_653[FLOAT, 16x16x1x1]
%onnx::Conv_656[FLOAT, 16x1x3x3]
%onnx::Conv_659[FLOAT, 16x16x1x1]
%onnx::Conv_662[FLOAT, 16x16x1x1]
%onnx::Conv_665[FLOAT, 16x1x3x3]
%onnx::Conv_668[FLOAT, 24x16x1x1]
%onnx::Conv_669[FLOAT, 24]
%onnx::Conv_671[FLOAT, 24x24x1x1]
%onnx::Conv_674[FLOAT, 24x1x3x3]
%onnx::Conv_677[FLOAT, 24x24x1x1]
%onnx::Conv_680[FLOAT, 144x24x1x1]
%onnx::Conv_681[FLOAT, 144]
%onnx::Conv_683[FLOAT, 144x1x5x5]
%onnx::Conv_686[FLOAT, 24x144x1x1]
%onnx::Conv_689[FLOAT, 24x12x1x1]
%onnx::Conv_692[FLOAT, 24x1x3x3]
%onnx::Conv_695[FLOAT, 24x12x1x1]
%onnx::Conv_698[FLOAT, 24x12x1x1]
%onnx::Conv_701[FLOAT, 24x1x5x5]
%onnx::Conv_704[FLOAT, 32x12x1x1]
%onnx::Conv_705[FLOAT, 32]
%onnx::Conv_707[FLOAT, 32x32x1x1]
%onnx::Conv_710[FLOAT, 32x1x5x5]
%onnx::Conv_713[FLOAT, 32x32x1x1]
%onnx::Conv_716[FLOAT, 32x16x1x1]
%onnx::Conv_719[FLOAT, 32x1x3x3]
%onnx::Conv_722[FLOAT, 32x16x1x1]
%onnx::Conv_725[FLOAT, 32x32x1x1]
%onnx::Conv_728[FLOAT, 32x1x5x5]
%onnx::Conv_731[FLOAT, 64x32x1x1]
%onnx::Conv_732[FLOAT, 64]
%onnx::Conv_734[FLOAT, 384x64x1x1]
%onnx::Conv_735[FLOAT, 384]
%onnx::Conv_737[FLOAT, 384x1x3x3]
%onnx::Conv_740[FLOAT, 64x384x1x1]
%onnx::Conv_743[FLOAT, 64x64x1x1]
%onnx::Conv_746[FLOAT, 64x1x3x3]
%onnx::Conv_749[FLOAT, 64x64x1x1]
%onnx::Conv_752[FLOAT, 64x64x1x1]
%onnx::Conv_755[FLOAT, 64x1x5x5]
%onnx::Conv_758[FLOAT, 64x64x1x1]
%onnx::Conv_761[FLOAT, 64x64x1x1]
%onnx::Conv_764[FLOAT, 64x1x5x5]
%onnx::Conv_767[FLOAT, 112x64x1x1]
%onnx::Conv_768[FLOAT, 112]
%onnx::Conv_770[FLOAT, 112x112x1x1]
%onnx::Conv_773[FLOAT, 112x1x3x3]
%onnx::Conv_776[FLOAT, 112x112x1x1]
%onnx::Conv_779[FLOAT, 112x56x1x1]
%onnx::Conv_782[FLOAT, 112x1x5x5]
%onnx::Conv_785[FLOAT, 112x56x1x1]
%onnx::Conv_788[FLOAT, 672x112x1x1]
%onnx::Conv_789[FLOAT, 672]
%onnx::Conv_791[FLOAT, 672x1x5x5]
%onnx::Conv_794[FLOAT, 112x672x1x1]
%onnx::Conv_797[FLOAT, 336x112x1x1]
%onnx::Conv_798[FLOAT, 336]
%onnx::Conv_800[FLOAT, 336x1x5x5]
%onnx::Conv_803[FLOAT, 184x336x1x1]
%onnx::Conv_804[FLOAT, 184]
%onnx::Conv_806[FLOAT, 184x184x1x1]
%onnx::Conv_809[FLOAT, 184x1x3x3]
%onnx::Conv_812[FLOAT, 184x184x1x1]
%onnx::Conv_815[FLOAT, 184x184x1x1]
%onnx::Conv_818[FLOAT, 184x1x3x3]
%onnx::Conv_821[FLOAT, 184x184x1x1]
%onnx::Conv_824[FLOAT, 184x184x1x1]
%onnx::Conv_827[FLOAT, 184x1x5x5]
%onnx::Conv_830[FLOAT, 184x184x1x1]
%onnx::Conv_833[FLOAT, 352x184x1x1]
%onnx::Conv_834[FLOAT, 352]
%onnx::Conv_836[FLOAT, 1504x352x1x1]
%onnx::Conv_837[FLOAT, 1504]
) {
%onnx::Conv_831 = Identity(%onnx::Conv_804)
%onnx::Conv_828 = Identity(%onnx::Conv_804)
%onnx::Conv_825 = Identity(%onnx::Conv_804)
%onnx::Conv_822 = Identity(%onnx::Conv_804)
%onnx::Conv_819 = Identity(%onnx::Conv_804)
%onnx::Conv_816 = Identity(%onnx::Conv_804)
%onnx::Conv_813 = Identity(%onnx::Conv_804)
%onnx::Conv_810 = Identity(%onnx::Conv_804)
%onnx::Conv_807 = Identity(%onnx::Conv_804)
%onnx::Conv_801 = Identity(%onnx::Conv_798)
%onnx::Conv_795 = Identity(%onnx::Conv_768)
%onnx::Conv_792 = Identity(%onnx::Conv_789)
%onnx::Conv_786 = Identity(%onnx::Conv_768)
%onnx::Conv_783 = Identity(%onnx::Conv_768)
%onnx::Conv_780 = Identity(%onnx::Conv_768)
%onnx::Conv_777 = Identity(%onnx::Conv_768)
%onnx::Conv_774 = Identity(%onnx::Conv_768)
%onnx::Conv_771 = Identity(%onnx::Conv_768)
%onnx::Conv_765 = Identity(%onnx::Conv_732)
%onnx::Conv_762 = Identity(%onnx::Conv_732)
%onnx::Conv_759 = Identity(%onnx::Conv_732)
%onnx::Conv_756 = Identity(%onnx::Conv_732)
%onnx::Conv_753 = Identity(%onnx::Conv_732)
%onnx::Conv_750 = Identity(%onnx::Conv_732)
%onnx::Conv_747 = Identity(%onnx::Conv_732)
%onnx::Conv_744 = Identity(%onnx::Conv_732)
%onnx::Conv_741 = Identity(%onnx::Conv_732)
%onnx::Conv_738 = Identity(%onnx::Conv_735)
%onnx::Conv_729 = Identity(%onnx::Conv_705)
%onnx::Conv_726 = Identity(%onnx::Conv_705)
%onnx::Conv_723 = Identity(%onnx::Conv_705)
%onnx::Conv_720 = Identity(%onnx::Conv_705)
%onnx::Conv_717 = Identity(%onnx::Conv_705)
%onnx::Conv_714 = Identity(%onnx::Conv_705)
%onnx::Conv_711 = Identity(%onnx::Conv_705)
%onnx::Conv_708 = Identity(%onnx::Conv_705)
%onnx::Conv_702 = Identity(%onnx::Conv_669)
%onnx::Conv_699 = Identity(%onnx::Conv_669)
%onnx::Conv_696 = Identity(%onnx::Conv_669)
%onnx::Conv_693 = Identity(%onnx::Conv_669)
%onnx::Conv_690 = Identity(%onnx::Conv_669)
%onnx::Conv_687 = Identity(%onnx::Conv_669)
%onnx::Conv_684 = Identity(%onnx::Conv_681)
%onnx::Conv_678 = Identity(%onnx::Conv_669)
%onnx::Conv_675 = Identity(%onnx::Conv_669)
%onnx::Conv_672 = Identity(%onnx::Conv_669)
%onnx::Conv_666 = Identity(%onnx::Conv_651)
%onnx::Conv_663 = Identity(%onnx::Conv_651)
%onnx::Conv_660 = Identity(%onnx::Conv_651)
%onnx::Conv_657 = Identity(%onnx::Conv_651)
%onnx::Conv_654 = Identity(%onnx::Conv_651)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_650, %onnx::Conv_651)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_803, %onnx::Conv_804)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_806, %onnx::Conv_807)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_809, %onnx::Conv_810)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_812, %onnx::Conv_813)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_815, %onnx::Conv_816)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_818, %onnx::Conv_819)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_821, %onnx::Conv_822)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_824, %onnx::Conv_825)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_827, %onnx::Conv_828)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_830, %onnx::Conv_831)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_833, %onnx::Conv_834)
%/cells.21/relu/Relu_output_0 = Relu(%/cells.21/conv/Conv_output_0)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/relu/Relu_output_0, %onnx::Conv_836, %onnx::Conv_837)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%648 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %648
}
|
val_accuracy
| 0
| 54,702,208
| 1,398,084
|
{'zcp_synflow': 79.74902368117297, 'zcp_zen': 65.44310760498047, 'zcp_epe_nas': 7.238702994174853, 'zcp_fisher': 0.0693884789943695, 'zcp_flops': 54702208.0, 'zcp_grad_norm': 18.171730041503906, 'zcp_grasp': -0.017963409423828125, 'zcp_jacov': -16.056430542693008, 'zcp_l2_norm': 564.806396484375, 'zcp_nwot': 208.49628042583248, 'zcp_params': 1398084.0, 'zcp_plain': 0.0022438233718276024, 'zcp_snip': 31.65863037109375, 'lat_1080ti_1': 0.5460402454039142, 'lat_1080ti_32': 0.5861421034334742, 'lat_1080ti_64': 0.4138129148335754, 'lat_2080ti_1': 0.6352319321004808, 'lat_2080ti_32': 0.6188877020105609, 'lat_2080ti_64': 0.4376905361732703, 'lat_essential_ph_1': 0.18867924528301888, 'lat_eyeriss': 0.27473273064016285, 'lat_fpga': 0.2578592703272345, 'lat_gold_6226': 0.1579714838457596, 'lat_gold_6240': 0.35782768681384447, 'lat_pixel2': 0.17391304347826086, 'lat_pixel3': 0.3128580231945633, 'lat_raspi4': 0.3105163814501688, 'lat_samsung_a50': 0.11578947368421053, 'lat_samsung_s7': 0.10236220472440945, 'lat_silver_4114': 0.37468853248124134, 'lat_silver_4210r': 0.41191612178854536, 'lat_titan_rtx_1': 0.6008550376958108, 'lat_titan_rtx_32': 0.6050202458388119, 'lat_titan_rtx_64': 0.49310075947742626, 'lat_titanx_1': 0.3178599027153094, 'lat_titanx_32': 0.5202532334978223, 'lat_titanx_64': 0.38817393954529916, 'lat_titanxp_1': 0.5929474368589497, 'lat_titanxp_32': 0.5598508771527165, 'lat_titanxp_64': 0.43465908872479503}
| |
FBNet_2054
|
FBNet
|
2054
|
2054
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_632[FLOAT, 16x3x3x3]
%onnx::Conv_633[FLOAT, 16]
%onnx::Conv_635[FLOAT, 48x16x1x1]
%onnx::Conv_636[FLOAT, 48]
%onnx::Conv_638[FLOAT, 48x1x5x5]
%onnx::Conv_641[FLOAT, 16x48x1x1]
%onnx::Conv_644[FLOAT, 48x16x1x1]
%onnx::Conv_647[FLOAT, 48x1x3x3]
%onnx::Conv_650[FLOAT, 24x48x1x1]
%onnx::Conv_651[FLOAT, 24]
%onnx::Conv_653[FLOAT, 24x12x1x1]
%onnx::Conv_656[FLOAT, 24x1x3x3]
%onnx::Conv_659[FLOAT, 24x12x1x1]
%onnx::Conv_662[FLOAT, 24x12x1x1]
%onnx::Conv_665[FLOAT, 24x1x3x3]
%onnx::Conv_668[FLOAT, 24x12x1x1]
%onnx::Conv_671[FLOAT, 24x12x1x1]
%onnx::Conv_674[FLOAT, 24x1x5x5]
%onnx::Conv_677[FLOAT, 32x12x1x1]
%onnx::Conv_678[FLOAT, 32]
%onnx::Conv_680[FLOAT, 32x32x1x1]
%onnx::Conv_683[FLOAT, 32x1x5x5]
%onnx::Conv_686[FLOAT, 32x32x1x1]
%onnx::Conv_689[FLOAT, 32x32x1x1]
%onnx::Conv_692[FLOAT, 32x1x5x5]
%onnx::Conv_695[FLOAT, 32x32x1x1]
%onnx::Conv_698[FLOAT, 192x32x1x1]
%onnx::Conv_699[FLOAT, 192]
%onnx::Conv_701[FLOAT, 192x1x3x3]
%onnx::Conv_704[FLOAT, 64x192x1x1]
%onnx::Conv_705[FLOAT, 64]
%onnx::Conv_707[FLOAT, 384x64x1x1]
%onnx::Conv_708[FLOAT, 384]
%onnx::Conv_710[FLOAT, 384x1x3x3]
%onnx::Conv_713[FLOAT, 64x384x1x1]
%onnx::Conv_716[FLOAT, 384x64x1x1]
%onnx::Conv_719[FLOAT, 384x1x3x3]
%onnx::Conv_722[FLOAT, 64x384x1x1]
%onnx::Conv_725[FLOAT, 192x64x1x1]
%onnx::Conv_728[FLOAT, 192x1x3x3]
%onnx::Conv_731[FLOAT, 64x192x1x1]
%onnx::Conv_734[FLOAT, 192x64x1x1]
%onnx::Conv_737[FLOAT, 192x1x3x3]
%onnx::Conv_740[FLOAT, 112x192x1x1]
%onnx::Conv_741[FLOAT, 112]
%onnx::Conv_743[FLOAT, 112x56x1x1]
%onnx::Conv_746[FLOAT, 112x1x3x3]
%onnx::Conv_749[FLOAT, 112x56x1x1]
%onnx::Conv_752[FLOAT, 336x112x1x1]
%onnx::Conv_753[FLOAT, 336]
%onnx::Conv_755[FLOAT, 336x1x3x3]
%onnx::Conv_758[FLOAT, 112x336x1x1]
%onnx::Conv_761[FLOAT, 336x112x1x1]
%onnx::Conv_764[FLOAT, 336x1x3x3]
%onnx::Conv_767[FLOAT, 112x336x1x1]
%onnx::Conv_770[FLOAT, 672x112x1x1]
%onnx::Conv_771[FLOAT, 672]
%onnx::Conv_773[FLOAT, 672x1x3x3]
%onnx::Conv_776[FLOAT, 184x672x1x1]
%onnx::Conv_777[FLOAT, 184]
%onnx::Conv_779[FLOAT, 184x92x1x1]
%onnx::Conv_782[FLOAT, 184x1x3x3]
%onnx::Conv_785[FLOAT, 184x92x1x1]
%onnx::Conv_788[FLOAT, 184x92x1x1]
%onnx::Conv_791[FLOAT, 184x1x3x3]
%onnx::Conv_794[FLOAT, 184x92x1x1]
%onnx::Conv_797[FLOAT, 352x184x1x1]
%onnx::Conv_798[FLOAT, 352]
%onnx::Conv_800[FLOAT, 1504x352x1x1]
%onnx::Conv_801[FLOAT, 1504]
) {
%onnx::Conv_795 = Identity(%onnx::Conv_777)
%onnx::Conv_792 = Identity(%onnx::Conv_777)
%onnx::Conv_789 = Identity(%onnx::Conv_777)
%onnx::Conv_786 = Identity(%onnx::Conv_777)
%onnx::Conv_783 = Identity(%onnx::Conv_777)
%onnx::Conv_780 = Identity(%onnx::Conv_777)
%onnx::Conv_774 = Identity(%onnx::Conv_771)
%onnx::Conv_768 = Identity(%onnx::Conv_741)
%onnx::Conv_765 = Identity(%onnx::Conv_753)
%onnx::Conv_762 = Identity(%onnx::Conv_753)
%onnx::Conv_759 = Identity(%onnx::Conv_741)
%onnx::Conv_756 = Identity(%onnx::Conv_753)
%onnx::Conv_750 = Identity(%onnx::Conv_741)
%onnx::Conv_747 = Identity(%onnx::Conv_741)
%onnx::Conv_744 = Identity(%onnx::Conv_741)
%onnx::Conv_738 = Identity(%onnx::Conv_699)
%onnx::Conv_735 = Identity(%onnx::Conv_699)
%onnx::Conv_732 = Identity(%onnx::Conv_705)
%onnx::Conv_729 = Identity(%onnx::Conv_699)
%onnx::Conv_726 = Identity(%onnx::Conv_699)
%onnx::Conv_723 = Identity(%onnx::Conv_705)
%onnx::Conv_720 = Identity(%onnx::Conv_708)
%onnx::Conv_717 = Identity(%onnx::Conv_708)
%onnx::Conv_714 = Identity(%onnx::Conv_705)
%onnx::Conv_711 = Identity(%onnx::Conv_708)
%onnx::Conv_702 = Identity(%onnx::Conv_699)
%onnx::Conv_696 = Identity(%onnx::Conv_678)
%onnx::Conv_693 = Identity(%onnx::Conv_678)
%onnx::Conv_690 = Identity(%onnx::Conv_678)
%onnx::Conv_687 = Identity(%onnx::Conv_678)
%onnx::Conv_684 = Identity(%onnx::Conv_678)
%onnx::Conv_681 = Identity(%onnx::Conv_678)
%onnx::Conv_675 = Identity(%onnx::Conv_651)
%onnx::Conv_672 = Identity(%onnx::Conv_651)
%onnx::Conv_669 = Identity(%onnx::Conv_651)
%onnx::Conv_666 = Identity(%onnx::Conv_651)
%onnx::Conv_663 = Identity(%onnx::Conv_651)
%onnx::Conv_660 = Identity(%onnx::Conv_651)
%onnx::Conv_657 = Identity(%onnx::Conv_651)
%onnx::Conv_654 = Identity(%onnx::Conv_651)
%onnx::Conv_648 = Identity(%onnx::Conv_636)
%onnx::Conv_645 = Identity(%onnx::Conv_636)
%onnx::Conv_642 = Identity(%onnx::Conv_633)
%onnx::Conv_639 = Identity(%onnx::Conv_636)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_632, %onnx::Conv_633)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_635, %onnx::Conv_636)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_638, %onnx::Conv_639)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_641, %onnx::Conv_642)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_644, %onnx::Conv_645)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_647, %onnx::Conv_648)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_650, %onnx::Conv_651)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_653, %onnx::Conv_654)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_656, %onnx::Conv_657)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_659, %onnx::Conv_660)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_662, %onnx::Conv_663)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_output_0 = Reshape(%/cells.3/nl/Relu_output_0, %/cells.3/shuffle/Constant_output_0)
%/cells.3/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.3/shuffle/Reshape_output_0)
%/cells.3/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.3/shuffle/Reshape_1_output_0 = Reshape(%/cells.3/shuffle/Transpose_output_0, %/cells.3/shuffle/Constant_1_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/shuffle/Reshape_1_output_0, %onnx::Conv_665, %onnx::Conv_666)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_668, %onnx::Conv_669)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_671, %onnx::Conv_672)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_674, %onnx::Conv_675)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_677, %onnx::Conv_678)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_680, %onnx::Conv_681)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_683, %onnx::Conv_684)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_686, %onnx::Conv_687)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_689, %onnx::Conv_690)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_692, %onnx::Conv_693)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_695, %onnx::Conv_696)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_698, %onnx::Conv_699)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_701, %onnx::Conv_702)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_704, %onnx::Conv_705)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_707, %onnx::Conv_708)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_710, %onnx::Conv_711)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_713, %onnx::Conv_714)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_716, %onnx::Conv_717)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_719, %onnx::Conv_720)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_722, %onnx::Conv_723)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_725, %onnx::Conv_726)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_728, %onnx::Conv_729)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_731, %onnx::Conv_732)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_734, %onnx::Conv_735)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_737, %onnx::Conv_738)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_740, %onnx::Conv_741)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_743, %onnx::Conv_744)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_746, %onnx::Conv_747)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_749, %onnx::Conv_750)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_752, %onnx::Conv_753)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_755, %onnx::Conv_756)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_758, %onnx::Conv_759)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_761, %onnx::Conv_762)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_764, %onnx::Conv_765)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_767, %onnx::Conv_768)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_770, %onnx::Conv_771)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_773, %onnx::Conv_774)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_776, %onnx::Conv_777)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_779, %onnx::Conv_780)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_782, %onnx::Conv_783)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_785, %onnx::Conv_786)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_788, %onnx::Conv_789)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_791, %onnx::Conv_792)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_794, %onnx::Conv_795)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_797, %onnx::Conv_798)
%/cells.21/relu/Relu_output_0 = Relu(%/cells.21/conv/Conv_output_0)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/relu/Relu_output_0, %onnx::Conv_800, %onnx::Conv_801)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%630 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %630
}
|
val_accuracy
| 0
| 53,302,400
| 1,412,284
|
{'zcp_synflow': 68.2574534340288, 'zcp_zen': 59.75991439819336, 'zcp_epe_nas': 7.80100153068974, 'zcp_fisher': 0.07742425799369812, 'zcp_flops': 53302400.0, 'zcp_grad_norm': 18.225624084472656, 'zcp_grasp': -0.09383010864257812, 'zcp_jacov': -16.057640525203652, 'zcp_l2_norm': 541.5895385742188, 'zcp_nwot': 207.57039576709042, 'zcp_params': 1412284.0, 'zcp_plain': 0.010103056207299232, 'zcp_snip': 29.85713005065918, 'lat_1080ti_1': 0.4229989381881782, 'lat_1080ti_32': 0.3898781314832276, 'lat_1080ti_64': 0.18994346140897597, 'lat_2080ti_1': 0.4833960799275503, 'lat_2080ti_32': 0.33768191421246735, 'lat_2080ti_64': 0.21400544081076486, 'lat_essential_ph_1': 0.20754716981132076, 'lat_eyeriss': 0.21946344415308158, 'lat_fpga': 0.236455931683472, 'lat_gold_6226': 0.25434289816513517, 'lat_gold_6240': 0.3823570934096024, 'lat_pixel2': 0.391304347826087, 'lat_pixel3': 0.21958237490270677, 'lat_raspi4': 0.21238828574854157, 'lat_samsung_a50': 0.10526315789473684, 'lat_samsung_s7': 0.11023622047244094, 'lat_silver_4114': 0.41015699885256446, 'lat_silver_4210r': 0.4641236994890747, 'lat_titan_rtx_1': 0.4452606340309013, 'lat_titan_rtx_32': 0.359500874405597, 'lat_titan_rtx_64': 0.23809075079245917, 'lat_titanx_1': 0.23692555396864923, 'lat_titanx_32': 0.2745638056182733, 'lat_titanx_64': 0.179271485145952, 'lat_titanxp_1': 0.41094379602643105, 'lat_titanxp_32': 0.3033319283194257, 'lat_titanxp_64': 0.19648421346222253}
| |
FBNet_1407
|
FBNet
|
1407
|
1407
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_660[FLOAT, 16x3x3x3]
%onnx::Conv_661[FLOAT, 16]
%onnx::Conv_663[FLOAT, 96x16x1x1]
%onnx::Conv_664[FLOAT, 96]
%onnx::Conv_666[FLOAT, 96x1x3x3]
%onnx::Conv_669[FLOAT, 16x96x1x1]
%onnx::Conv_672[FLOAT, 96x16x1x1]
%onnx::Conv_675[FLOAT, 96x1x5x5]
%onnx::Conv_678[FLOAT, 24x96x1x1]
%onnx::Conv_679[FLOAT, 24]
%onnx::Conv_681[FLOAT, 24x12x1x1]
%onnx::Conv_684[FLOAT, 24x1x5x5]
%onnx::Conv_687[FLOAT, 24x12x1x1]
%onnx::Conv_690[FLOAT, 72x24x1x1]
%onnx::Conv_691[FLOAT, 72]
%onnx::Conv_693[FLOAT, 72x1x3x3]
%onnx::Conv_696[FLOAT, 24x72x1x1]
%onnx::Conv_699[FLOAT, 24x12x1x1]
%onnx::Conv_702[FLOAT, 24x1x5x5]
%onnx::Conv_705[FLOAT, 24x12x1x1]
%onnx::Conv_708[FLOAT, 24x24x1x1]
%onnx::Conv_711[FLOAT, 24x1x3x3]
%onnx::Conv_714[FLOAT, 32x24x1x1]
%onnx::Conv_715[FLOAT, 32]
%onnx::Conv_717[FLOAT, 32x16x1x1]
%onnx::Conv_720[FLOAT, 32x1x5x5]
%onnx::Conv_723[FLOAT, 32x16x1x1]
%onnx::Conv_726[FLOAT, 32x32x1x1]
%onnx::Conv_729[FLOAT, 32x1x3x3]
%onnx::Conv_732[FLOAT, 32x32x1x1]
%onnx::Conv_735[FLOAT, 32x32x1x1]
%onnx::Conv_738[FLOAT, 32x1x3x3]
%onnx::Conv_741[FLOAT, 32x32x1x1]
%onnx::Conv_744[FLOAT, 192x32x1x1]
%onnx::Conv_745[FLOAT, 192]
%onnx::Conv_747[FLOAT, 192x1x3x3]
%onnx::Conv_750[FLOAT, 64x192x1x1]
%onnx::Conv_751[FLOAT, 64]
%onnx::Conv_753[FLOAT, 64x64x1x1]
%onnx::Conv_756[FLOAT, 64x1x3x3]
%onnx::Conv_759[FLOAT, 64x64x1x1]
%onnx::Conv_762[FLOAT, 64x32x1x1]
%onnx::Conv_765[FLOAT, 64x1x5x5]
%onnx::Conv_768[FLOAT, 64x32x1x1]
%onnx::Conv_771[FLOAT, 64x32x1x1]
%onnx::Conv_774[FLOAT, 64x1x3x3]
%onnx::Conv_777[FLOAT, 112x32x1x1]
%onnx::Conv_778[FLOAT, 112]
%onnx::Conv_780[FLOAT, 336x112x1x1]
%onnx::Conv_781[FLOAT, 336]
%onnx::Conv_783[FLOAT, 336x1x5x5]
%onnx::Conv_786[FLOAT, 112x336x1x1]
%onnx::Conv_789[FLOAT, 336x112x1x1]
%onnx::Conv_792[FLOAT, 336x1x3x3]
%onnx::Conv_795[FLOAT, 112x336x1x1]
%onnx::Conv_798[FLOAT, 672x112x1x1]
%onnx::Conv_799[FLOAT, 672]
%onnx::Conv_801[FLOAT, 672x1x5x5]
%onnx::Conv_804[FLOAT, 112x672x1x1]
%onnx::Conv_807[FLOAT, 112x56x1x1]
%onnx::Conv_810[FLOAT, 112x1x3x3]
%onnx::Conv_813[FLOAT, 184x56x1x1]
%onnx::Conv_814[FLOAT, 184]
%onnx::Conv_816[FLOAT, 552x184x1x1]
%onnx::Conv_817[FLOAT, 552]
%onnx::Conv_819[FLOAT, 552x1x3x3]
%onnx::Conv_822[FLOAT, 184x552x1x1]
%onnx::Conv_825[FLOAT, 552x184x1x1]
%onnx::Conv_828[FLOAT, 552x1x3x3]
%onnx::Conv_831[FLOAT, 184x552x1x1]
%onnx::Conv_834[FLOAT, 352x184x1x1]
%onnx::Conv_835[FLOAT, 352]
%onnx::Conv_837[FLOAT, 1504x352x1x1]
%onnx::Conv_838[FLOAT, 1504]
) {
%onnx::Conv_832 = Identity(%onnx::Conv_814)
%onnx::Conv_829 = Identity(%onnx::Conv_817)
%onnx::Conv_826 = Identity(%onnx::Conv_817)
%onnx::Conv_823 = Identity(%onnx::Conv_814)
%onnx::Conv_820 = Identity(%onnx::Conv_817)
%onnx::Conv_811 = Identity(%onnx::Conv_778)
%onnx::Conv_808 = Identity(%onnx::Conv_778)
%onnx::Conv_805 = Identity(%onnx::Conv_778)
%onnx::Conv_802 = Identity(%onnx::Conv_799)
%onnx::Conv_796 = Identity(%onnx::Conv_778)
%onnx::Conv_793 = Identity(%onnx::Conv_781)
%onnx::Conv_790 = Identity(%onnx::Conv_781)
%onnx::Conv_787 = Identity(%onnx::Conv_778)
%onnx::Conv_784 = Identity(%onnx::Conv_781)
%onnx::Conv_775 = Identity(%onnx::Conv_751)
%onnx::Conv_772 = Identity(%onnx::Conv_751)
%onnx::Conv_769 = Identity(%onnx::Conv_751)
%onnx::Conv_766 = Identity(%onnx::Conv_751)
%onnx::Conv_763 = Identity(%onnx::Conv_751)
%onnx::Conv_760 = Identity(%onnx::Conv_751)
%onnx::Conv_757 = Identity(%onnx::Conv_751)
%onnx::Conv_754 = Identity(%onnx::Conv_751)
%onnx::Conv_748 = Identity(%onnx::Conv_745)
%onnx::Conv_742 = Identity(%onnx::Conv_715)
%onnx::Conv_739 = Identity(%onnx::Conv_715)
%onnx::Conv_736 = Identity(%onnx::Conv_715)
%onnx::Conv_733 = Identity(%onnx::Conv_715)
%onnx::Conv_730 = Identity(%onnx::Conv_715)
%onnx::Conv_727 = Identity(%onnx::Conv_715)
%onnx::Conv_724 = Identity(%onnx::Conv_715)
%onnx::Conv_721 = Identity(%onnx::Conv_715)
%onnx::Conv_718 = Identity(%onnx::Conv_715)
%onnx::Conv_712 = Identity(%onnx::Conv_679)
%onnx::Conv_709 = Identity(%onnx::Conv_679)
%onnx::Conv_706 = Identity(%onnx::Conv_679)
%onnx::Conv_703 = Identity(%onnx::Conv_679)
%onnx::Conv_700 = Identity(%onnx::Conv_679)
%onnx::Conv_697 = Identity(%onnx::Conv_679)
%onnx::Conv_694 = Identity(%onnx::Conv_691)
%onnx::Conv_688 = Identity(%onnx::Conv_679)
%onnx::Conv_685 = Identity(%onnx::Conv_679)
%onnx::Conv_682 = Identity(%onnx::Conv_679)
%onnx::Conv_676 = Identity(%onnx::Conv_664)
%onnx::Conv_673 = Identity(%onnx::Conv_664)
%onnx::Conv_670 = Identity(%onnx::Conv_661)
%onnx::Conv_667 = Identity(%onnx::Conv_664)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_660, %onnx::Conv_661)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.21/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.21/relu/Relu_output_0 = Relu(%/cells.21/conv/Conv_output_0)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/relu/Relu_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%658 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %658
}
|
val_accuracy
| 0
| 64,823,424
| 1,593,812
|
{'zcp_synflow': 71.96635132034868, 'zcp_zen': 61.743160247802734, 'zcp_epe_nas': 8.608805977597298, 'zcp_fisher': 0.11580222845077515, 'zcp_flops': 64823424.0, 'zcp_grad_norm': 20.94672966003418, 'zcp_grasp': -0.0236968994140625, 'zcp_jacov': -16.058393459197234, 'zcp_l2_norm': 552.0489501953125, 'zcp_nwot': 212.79490796255152, 'zcp_params': 1593812.0, 'zcp_plain': 0.0028113939333707094, 'zcp_snip': 35.34382247924805, 'lat_1080ti_1': 0.5591726545428107, 'lat_1080ti_32': 0.513836753807461, 'lat_1080ti_64': 0.4642613143171164, 'lat_2080ti_1': 0.5734417879148771, 'lat_2080ti_32': 0.5255690106954026, 'lat_2080ti_64': 0.4712260239463605, 'lat_essential_ph_1': 0.20754716981132076, 'lat_eyeriss': 0.37533236755725874, 'lat_fpga': 0.41016633847554396, 'lat_gold_6226': 0.27739757626738265, 'lat_gold_6240': 0.4467312730713451, 'lat_pixel2': 0.2391304347826087, 'lat_pixel3': 0.4203410593004545, 'lat_raspi4': 0.38602801512432877, 'lat_samsung_a50': 0.15789473684210525, 'lat_samsung_s7': 0.1889763779527559, 'lat_silver_4114': 0.4836811156659293, 'lat_silver_4210r': 0.5062041943734915, 'lat_titan_rtx_1': 0.5459494421368645, 'lat_titan_rtx_32': 0.49358805372280135, 'lat_titan_rtx_64': 0.48459026252261006, 'lat_titanx_1': 0.289312218717299, 'lat_titanx_32': 0.4868110083394912, 'lat_titanx_64': 0.4373934883160455, 'lat_titanxp_1': 0.520653071878602, 'lat_titanxp_32': 0.5024042054678501, 'lat_titanxp_64': 0.4864367780915869}
| |
FBNet_2073
|
FBNet
|
2073
|
2073
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_613[FLOAT, 16x3x3x3]
%onnx::Conv_614[FLOAT, 16]
%onnx::Conv_616[FLOAT, 16x16x1x1]
%onnx::Conv_619[FLOAT, 16x1x5x5]
%onnx::Conv_622[FLOAT, 16x16x1x1]
%onnx::Conv_625[FLOAT, 48x16x1x1]
%onnx::Conv_626[FLOAT, 48]
%onnx::Conv_628[FLOAT, 48x1x3x3]
%onnx::Conv_631[FLOAT, 24x48x1x1]
%onnx::Conv_632[FLOAT, 24]
%onnx::Conv_634[FLOAT, 144x24x1x1]
%onnx::Conv_635[FLOAT, 144]
%onnx::Conv_637[FLOAT, 144x1x5x5]
%onnx::Conv_640[FLOAT, 24x144x1x1]
%onnx::Conv_643[FLOAT, 72x24x1x1]
%onnx::Conv_644[FLOAT, 72]
%onnx::Conv_646[FLOAT, 72x1x5x5]
%onnx::Conv_649[FLOAT, 24x72x1x1]
%onnx::Conv_652[FLOAT, 24x12x1x1]
%onnx::Conv_655[FLOAT, 24x1x5x5]
%onnx::Conv_658[FLOAT, 24x12x1x1]
%onnx::Conv_661[FLOAT, 72x24x1x1]
%onnx::Conv_664[FLOAT, 72x1x3x3]
%onnx::Conv_667[FLOAT, 32x72x1x1]
%onnx::Conv_668[FLOAT, 32]
%onnx::Conv_670[FLOAT, 96x32x1x1]
%onnx::Conv_671[FLOAT, 96]
%onnx::Conv_673[FLOAT, 96x1x5x5]
%onnx::Conv_676[FLOAT, 32x96x1x1]
%onnx::Conv_679[FLOAT, 32x32x1x1]
%onnx::Conv_682[FLOAT, 32x1x5x5]
%onnx::Conv_685[FLOAT, 32x32x1x1]
%onnx::Conv_688[FLOAT, 32x32x1x1]
%onnx::Conv_691[FLOAT, 32x1x5x5]
%onnx::Conv_694[FLOAT, 64x32x1x1]
%onnx::Conv_695[FLOAT, 64]
%onnx::Conv_697[FLOAT, 192x64x1x1]
%onnx::Conv_698[FLOAT, 192]
%onnx::Conv_700[FLOAT, 192x1x3x3]
%onnx::Conv_703[FLOAT, 64x192x1x1]
%onnx::Conv_706[FLOAT, 384x64x1x1]
%onnx::Conv_707[FLOAT, 384]
%onnx::Conv_709[FLOAT, 384x1x5x5]
%onnx::Conv_712[FLOAT, 64x384x1x1]
%onnx::Conv_715[FLOAT, 64x64x1x1]
%onnx::Conv_718[FLOAT, 64x1x5x5]
%onnx::Conv_721[FLOAT, 112x64x1x1]
%onnx::Conv_722[FLOAT, 112]
%onnx::Conv_724[FLOAT, 112x56x1x1]
%onnx::Conv_727[FLOAT, 112x1x5x5]
%onnx::Conv_730[FLOAT, 112x56x1x1]
%onnx::Conv_733[FLOAT, 112x112x1x1]
%onnx::Conv_736[FLOAT, 112x1x5x5]
%onnx::Conv_739[FLOAT, 112x112x1x1]
%onnx::Conv_742[FLOAT, 336x112x1x1]
%onnx::Conv_743[FLOAT, 336]
%onnx::Conv_745[FLOAT, 336x1x3x3]
%onnx::Conv_748[FLOAT, 112x336x1x1]
%onnx::Conv_751[FLOAT, 112x56x1x1]
%onnx::Conv_754[FLOAT, 112x1x3x3]
%onnx::Conv_757[FLOAT, 184x56x1x1]
%onnx::Conv_758[FLOAT, 184]
%onnx::Conv_760[FLOAT, 552x184x1x1]
%onnx::Conv_761[FLOAT, 552]
%onnx::Conv_763[FLOAT, 552x1x5x5]
%onnx::Conv_766[FLOAT, 184x552x1x1]
%onnx::Conv_769[FLOAT, 184x92x1x1]
%onnx::Conv_772[FLOAT, 184x1x5x5]
%onnx::Conv_775[FLOAT, 184x92x1x1]
%onnx::Conv_778[FLOAT, 184x184x1x1]
%onnx::Conv_781[FLOAT, 184x1x5x5]
%onnx::Conv_784[FLOAT, 352x184x1x1]
%onnx::Conv_785[FLOAT, 352]
%onnx::Conv_787[FLOAT, 1504x352x1x1]
%onnx::Conv_788[FLOAT, 1504]
) {
%onnx::Conv_782 = Identity(%onnx::Conv_758)
%onnx::Conv_779 = Identity(%onnx::Conv_758)
%onnx::Conv_776 = Identity(%onnx::Conv_758)
%onnx::Conv_773 = Identity(%onnx::Conv_758)
%onnx::Conv_770 = Identity(%onnx::Conv_758)
%onnx::Conv_767 = Identity(%onnx::Conv_758)
%onnx::Conv_764 = Identity(%onnx::Conv_761)
%onnx::Conv_755 = Identity(%onnx::Conv_722)
%onnx::Conv_752 = Identity(%onnx::Conv_722)
%onnx::Conv_749 = Identity(%onnx::Conv_722)
%onnx::Conv_746 = Identity(%onnx::Conv_743)
%onnx::Conv_740 = Identity(%onnx::Conv_722)
%onnx::Conv_737 = Identity(%onnx::Conv_722)
%onnx::Conv_734 = Identity(%onnx::Conv_722)
%onnx::Conv_731 = Identity(%onnx::Conv_722)
%onnx::Conv_728 = Identity(%onnx::Conv_722)
%onnx::Conv_725 = Identity(%onnx::Conv_722)
%onnx::Conv_719 = Identity(%onnx::Conv_695)
%onnx::Conv_716 = Identity(%onnx::Conv_695)
%onnx::Conv_713 = Identity(%onnx::Conv_695)
%onnx::Conv_710 = Identity(%onnx::Conv_707)
%onnx::Conv_704 = Identity(%onnx::Conv_695)
%onnx::Conv_701 = Identity(%onnx::Conv_698)
%onnx::Conv_692 = Identity(%onnx::Conv_668)
%onnx::Conv_689 = Identity(%onnx::Conv_668)
%onnx::Conv_686 = Identity(%onnx::Conv_668)
%onnx::Conv_683 = Identity(%onnx::Conv_668)
%onnx::Conv_680 = Identity(%onnx::Conv_668)
%onnx::Conv_677 = Identity(%onnx::Conv_668)
%onnx::Conv_674 = Identity(%onnx::Conv_671)
%onnx::Conv_665 = Identity(%onnx::Conv_644)
%onnx::Conv_662 = Identity(%onnx::Conv_644)
%onnx::Conv_659 = Identity(%onnx::Conv_632)
%onnx::Conv_656 = Identity(%onnx::Conv_632)
%onnx::Conv_653 = Identity(%onnx::Conv_632)
%onnx::Conv_650 = Identity(%onnx::Conv_632)
%onnx::Conv_647 = Identity(%onnx::Conv_644)
%onnx::Conv_641 = Identity(%onnx::Conv_632)
%onnx::Conv_638 = Identity(%onnx::Conv_635)
%onnx::Conv_629 = Identity(%onnx::Conv_626)
%onnx::Conv_623 = Identity(%onnx::Conv_614)
%onnx::Conv_620 = Identity(%onnx::Conv_614)
%onnx::Conv_617 = Identity(%onnx::Conv_614)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_613, %onnx::Conv_614)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_616, %onnx::Conv_617)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_619, %onnx::Conv_620)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_622, %onnx::Conv_623)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_625, %onnx::Conv_626)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_628, %onnx::Conv_629)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_631, %onnx::Conv_632)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_634, %onnx::Conv_635)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_637, %onnx::Conv_638)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_640, %onnx::Conv_641)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_643, %onnx::Conv_644)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_646, %onnx::Conv_647)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_649, %onnx::Conv_650)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_652, %onnx::Conv_653)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_655, %onnx::Conv_656)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_658, %onnx::Conv_659)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%611 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %611
}
|
val_accuracy
| 0
| 58,964,352
| 1,333,748
|
{'zcp_synflow': 74.57992752721901, 'zcp_zen': 63.14750671386719, 'zcp_epe_nas': 5.742881528725499, 'zcp_fisher': 0.10832878202199936, 'zcp_flops': 58964352.0, 'zcp_grad_norm': 22.995986938476562, 'zcp_grasp': 0.047260284423828125, 'zcp_jacov': -16.06291595742394, 'zcp_l2_norm': 532.3233642578125, 'zcp_nwot': 211.9631139867734, 'zcp_params': 1333748.0, 'zcp_plain': 0.009027144871652126, 'zcp_snip': 36.62528991699219, 'lat_1080ti_1': 0.4651189521850695, 'lat_1080ti_32': 0.5020710672197949, 'lat_1080ti_64': 0.4596454689025721, 'lat_2080ti_1': 0.4856556579918911, 'lat_2080ti_32': 0.5241692338643543, 'lat_2080ti_64': 0.44611264449914007, 'lat_essential_ph_1': 0.20754716981132076, 'lat_eyeriss': 0.35543615792272576, 'lat_fpga': 0.307830079458534, 'lat_gold_6226': 0.18252163613230385, 'lat_gold_6240': 0.2799394314487959, 'lat_pixel2': 0.17391304347826086, 'lat_pixel3': 0.4081354358494352, 'lat_raspi4': 0.38597302224542557, 'lat_samsung_a50': 0.12631578947368421, 'lat_samsung_s7': 0.10236220472440945, 'lat_silver_4114': 0.33080758546489897, 'lat_silver_4210r': 0.29331376995555336, 'lat_titan_rtx_1': 0.46969717872291505, 'lat_titan_rtx_32': 0.5116606976372339, 'lat_titan_rtx_64': 0.4556244464630431, 'lat_titanx_1': 0.23988020926942474, 'lat_titanx_32': 0.48245666868953024, 'lat_titanx_64': 0.44624826819232516, 'lat_titanxp_1': 0.42508223473742573, 'lat_titanxp_32': 0.5146697488423926, 'lat_titanxp_64': 0.4828225702659326}
| |
FBNet_3127
|
FBNet
|
3127
|
3127
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_603[FLOAT, 16x3x3x3]
%onnx::Conv_604[FLOAT, 16]
%onnx::Conv_606[FLOAT, 16x16x1x1]
%onnx::Conv_609[FLOAT, 16x1x5x5]
%onnx::Conv_612[FLOAT, 16x16x1x1]
%onnx::Conv_615[FLOAT, 48x16x1x1]
%onnx::Conv_616[FLOAT, 48]
%onnx::Conv_618[FLOAT, 48x1x3x3]
%onnx::Conv_621[FLOAT, 24x48x1x1]
%onnx::Conv_622[FLOAT, 24]
%onnx::Conv_624[FLOAT, 144x24x1x1]
%onnx::Conv_625[FLOAT, 144]
%onnx::Conv_627[FLOAT, 144x1x5x5]
%onnx::Conv_630[FLOAT, 24x144x1x1]
%onnx::Conv_633[FLOAT, 24x24x1x1]
%onnx::Conv_636[FLOAT, 24x1x3x3]
%onnx::Conv_639[FLOAT, 24x24x1x1]
%onnx::Conv_642[FLOAT, 144x24x1x1]
%onnx::Conv_645[FLOAT, 144x1x5x5]
%onnx::Conv_648[FLOAT, 32x144x1x1]
%onnx::Conv_649[FLOAT, 32]
%onnx::Conv_651[FLOAT, 32x16x1x1]
%onnx::Conv_654[FLOAT, 32x1x3x3]
%onnx::Conv_657[FLOAT, 32x16x1x1]
%onnx::Conv_660[FLOAT, 96x32x1x1]
%onnx::Conv_661[FLOAT, 96]
%onnx::Conv_663[FLOAT, 96x1x5x5]
%onnx::Conv_666[FLOAT, 32x96x1x1]
%onnx::Conv_669[FLOAT, 192x32x1x1]
%onnx::Conv_670[FLOAT, 192]
%onnx::Conv_672[FLOAT, 192x1x5x5]
%onnx::Conv_675[FLOAT, 64x192x1x1]
%onnx::Conv_676[FLOAT, 64]
%onnx::Conv_678[FLOAT, 192x64x1x1]
%onnx::Conv_681[FLOAT, 192x1x3x3]
%onnx::Conv_684[FLOAT, 64x192x1x1]
%onnx::Conv_687[FLOAT, 64x32x1x1]
%onnx::Conv_690[FLOAT, 64x1x5x5]
%onnx::Conv_693[FLOAT, 64x32x1x1]
%onnx::Conv_696[FLOAT, 64x64x1x1]
%onnx::Conv_699[FLOAT, 64x1x5x5]
%onnx::Conv_702[FLOAT, 64x64x1x1]
%onnx::Conv_705[FLOAT, 64x64x1x1]
%onnx::Conv_708[FLOAT, 64x1x5x5]
%onnx::Conv_711[FLOAT, 112x64x1x1]
%onnx::Conv_712[FLOAT, 112]
%onnx::Conv_714[FLOAT, 336x112x1x1]
%onnx::Conv_715[FLOAT, 336]
%onnx::Conv_717[FLOAT, 336x1x5x5]
%onnx::Conv_720[FLOAT, 112x336x1x1]
%onnx::Conv_723[FLOAT, 112x112x1x1]
%onnx::Conv_726[FLOAT, 112x1x3x3]
%onnx::Conv_729[FLOAT, 112x112x1x1]
%onnx::Conv_732[FLOAT, 336x112x1x1]
%onnx::Conv_735[FLOAT, 336x1x5x5]
%onnx::Conv_738[FLOAT, 112x336x1x1]
%onnx::Conv_741[FLOAT, 672x112x1x1]
%onnx::Conv_742[FLOAT, 672]
%onnx::Conv_744[FLOAT, 672x1x5x5]
%onnx::Conv_747[FLOAT, 184x672x1x1]
%onnx::Conv_748[FLOAT, 184]
%onnx::Conv_750[FLOAT, 1104x184x1x1]
%onnx::Conv_751[FLOAT, 1104]
%onnx::Conv_753[FLOAT, 1104x1x5x5]
%onnx::Conv_756[FLOAT, 184x1104x1x1]
%onnx::Conv_759[FLOAT, 1104x184x1x1]
%onnx::Conv_762[FLOAT, 1104x1x3x3]
%onnx::Conv_765[FLOAT, 184x1104x1x1]
%onnx::Conv_768[FLOAT, 184x184x1x1]
%onnx::Conv_771[FLOAT, 184x1x5x5]
%onnx::Conv_774[FLOAT, 184x184x1x1]
%onnx::Conv_777[FLOAT, 184x184x1x1]
%onnx::Conv_780[FLOAT, 184x1x3x3]
%onnx::Conv_783[FLOAT, 352x184x1x1]
%onnx::Conv_784[FLOAT, 352]
%onnx::Conv_786[FLOAT, 1504x352x1x1]
%onnx::Conv_787[FLOAT, 1504]
) {
%onnx::Conv_781 = Identity(%onnx::Conv_748)
%onnx::Conv_778 = Identity(%onnx::Conv_748)
%onnx::Conv_775 = Identity(%onnx::Conv_748)
%onnx::Conv_772 = Identity(%onnx::Conv_748)
%onnx::Conv_769 = Identity(%onnx::Conv_748)
%onnx::Conv_766 = Identity(%onnx::Conv_748)
%onnx::Conv_763 = Identity(%onnx::Conv_751)
%onnx::Conv_760 = Identity(%onnx::Conv_751)
%onnx::Conv_757 = Identity(%onnx::Conv_748)
%onnx::Conv_754 = Identity(%onnx::Conv_751)
%onnx::Conv_745 = Identity(%onnx::Conv_742)
%onnx::Conv_739 = Identity(%onnx::Conv_712)
%onnx::Conv_736 = Identity(%onnx::Conv_715)
%onnx::Conv_733 = Identity(%onnx::Conv_715)
%onnx::Conv_730 = Identity(%onnx::Conv_712)
%onnx::Conv_727 = Identity(%onnx::Conv_712)
%onnx::Conv_724 = Identity(%onnx::Conv_712)
%onnx::Conv_721 = Identity(%onnx::Conv_712)
%onnx::Conv_718 = Identity(%onnx::Conv_715)
%onnx::Conv_709 = Identity(%onnx::Conv_676)
%onnx::Conv_706 = Identity(%onnx::Conv_676)
%onnx::Conv_703 = Identity(%onnx::Conv_676)
%onnx::Conv_700 = Identity(%onnx::Conv_676)
%onnx::Conv_697 = Identity(%onnx::Conv_676)
%onnx::Conv_694 = Identity(%onnx::Conv_676)
%onnx::Conv_691 = Identity(%onnx::Conv_676)
%onnx::Conv_688 = Identity(%onnx::Conv_676)
%onnx::Conv_685 = Identity(%onnx::Conv_676)
%onnx::Conv_682 = Identity(%onnx::Conv_670)
%onnx::Conv_679 = Identity(%onnx::Conv_670)
%onnx::Conv_673 = Identity(%onnx::Conv_670)
%onnx::Conv_667 = Identity(%onnx::Conv_649)
%onnx::Conv_664 = Identity(%onnx::Conv_661)
%onnx::Conv_658 = Identity(%onnx::Conv_649)
%onnx::Conv_655 = Identity(%onnx::Conv_649)
%onnx::Conv_652 = Identity(%onnx::Conv_649)
%onnx::Conv_646 = Identity(%onnx::Conv_625)
%onnx::Conv_643 = Identity(%onnx::Conv_625)
%onnx::Conv_640 = Identity(%onnx::Conv_622)
%onnx::Conv_637 = Identity(%onnx::Conv_622)
%onnx::Conv_634 = Identity(%onnx::Conv_622)
%onnx::Conv_631 = Identity(%onnx::Conv_622)
%onnx::Conv_628 = Identity(%onnx::Conv_625)
%onnx::Conv_619 = Identity(%onnx::Conv_616)
%onnx::Conv_613 = Identity(%onnx::Conv_604)
%onnx::Conv_610 = Identity(%onnx::Conv_604)
%onnx::Conv_607 = Identity(%onnx::Conv_604)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_603, %onnx::Conv_604)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_606, %onnx::Conv_607)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_609, %onnx::Conv_610)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_612, %onnx::Conv_613)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_615, %onnx::Conv_616)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_618, %onnx::Conv_619)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_621, %onnx::Conv_622)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_624, %onnx::Conv_625)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_627, %onnx::Conv_628)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_630, %onnx::Conv_631)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%601 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %601
}
|
val_accuracy
| 0
| 77,794,688
| 2,254,140
|
{'zcp_synflow': 81.20328913839799, 'zcp_zen': 71.361083984375, 'zcp_epe_nas': 9.87551809083078, 'zcp_fisher': 0.11348476260900497, 'zcp_flops': 77794688.0, 'zcp_grad_norm': 26.013874053955078, 'zcp_grasp': -0.010517120361328125, 'zcp_jacov': -16.0461662777636, 'zcp_l2_norm': 670.0664672851562, 'zcp_nwot': 213.43906015358496, 'zcp_params': 2254140.0, 'zcp_plain': 0.0010176384821534157, 'zcp_snip': 44.72586441040039, 'lat_1080ti_1': 0.539534304198715, 'lat_1080ti_32': 0.5680968486880984, 'lat_1080ti_64': 0.47799406459123056, 'lat_2080ti_1': 0.5590260364014887, 'lat_2080ti_32': 0.5399832022840166, 'lat_2080ti_64': 0.46778374036714004, 'lat_essential_ph_1': 0.37735849056603776, 'lat_eyeriss': 0.5513267196009756, 'lat_fpga': 0.5136702322405723, 'lat_gold_6226': 0.46081427249527684, 'lat_gold_6240': 0.5274585325866178, 'lat_pixel2': 0.3695652173913043, 'lat_pixel3': 0.5374234005834247, 'lat_raspi4': 0.5699029134439217, 'lat_samsung_a50': 0.23157894736842105, 'lat_samsung_s7': 0.2047244094488189, 'lat_silver_4114': 0.5327469155866936, 'lat_silver_4210r': 0.5383366385175075, 'lat_titan_rtx_1': 0.5319566826836866, 'lat_titan_rtx_32': 0.5222619322167428, 'lat_titan_rtx_64': 0.48703364715328595, 'lat_titanx_1': 0.2882708749108645, 'lat_titanx_32': 0.5377394142041394, 'lat_titanx_64': 0.47757018531035866, 'lat_titanxp_1': 0.5141107609269483, 'lat_titanxp_32': 0.5386296641098375, 'lat_titanxp_64': 0.48221226012117474}
| |
FBNet_4201
|
FBNet
|
4201
|
4201
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_630[FLOAT, 16x3x3x3]
%onnx::Conv_631[FLOAT, 16]
%onnx::Conv_633[FLOAT, 16x8x1x1]
%onnx::Conv_636[FLOAT, 16x1x5x5]
%onnx::Conv_639[FLOAT, 16x8x1x1]
%onnx::Conv_642[FLOAT, 96x16x1x1]
%onnx::Conv_643[FLOAT, 96]
%onnx::Conv_645[FLOAT, 96x1x3x3]
%onnx::Conv_648[FLOAT, 24x96x1x1]
%onnx::Conv_649[FLOAT, 24]
%onnx::Conv_651[FLOAT, 144x24x1x1]
%onnx::Conv_652[FLOAT, 144]
%onnx::Conv_654[FLOAT, 144x1x3x3]
%onnx::Conv_657[FLOAT, 24x144x1x1]
%onnx::Conv_660[FLOAT, 24x24x1x1]
%onnx::Conv_663[FLOAT, 24x1x3x3]
%onnx::Conv_666[FLOAT, 24x24x1x1]
%onnx::Conv_669[FLOAT, 144x24x1x1]
%onnx::Conv_672[FLOAT, 144x1x3x3]
%onnx::Conv_675[FLOAT, 32x144x1x1]
%onnx::Conv_676[FLOAT, 32]
%onnx::Conv_678[FLOAT, 32x32x1x1]
%onnx::Conv_681[FLOAT, 32x1x5x5]
%onnx::Conv_684[FLOAT, 32x32x1x1]
%onnx::Conv_687[FLOAT, 192x32x1x1]
%onnx::Conv_688[FLOAT, 192]
%onnx::Conv_690[FLOAT, 192x1x3x3]
%onnx::Conv_693[FLOAT, 32x192x1x1]
%onnx::Conv_696[FLOAT, 96x32x1x1]
%onnx::Conv_699[FLOAT, 96x1x5x5]
%onnx::Conv_702[FLOAT, 32x96x1x1]
%onnx::Conv_705[FLOAT, 192x32x1x1]
%onnx::Conv_708[FLOAT, 192x1x3x3]
%onnx::Conv_711[FLOAT, 64x192x1x1]
%onnx::Conv_712[FLOAT, 64]
%onnx::Conv_714[FLOAT, 64x64x1x1]
%onnx::Conv_717[FLOAT, 64x1x5x5]
%onnx::Conv_720[FLOAT, 64x64x1x1]
%onnx::Conv_723[FLOAT, 384x64x1x1]
%onnx::Conv_724[FLOAT, 384]
%onnx::Conv_726[FLOAT, 384x1x5x5]
%onnx::Conv_729[FLOAT, 64x384x1x1]
%onnx::Conv_732[FLOAT, 192x64x1x1]
%onnx::Conv_735[FLOAT, 192x1x5x5]
%onnx::Conv_738[FLOAT, 64x192x1x1]
%onnx::Conv_741[FLOAT, 64x64x1x1]
%onnx::Conv_744[FLOAT, 64x1x5x5]
%onnx::Conv_747[FLOAT, 112x64x1x1]
%onnx::Conv_748[FLOAT, 112]
%onnx::Conv_750[FLOAT, 672x112x1x1]
%onnx::Conv_751[FLOAT, 672]
%onnx::Conv_753[FLOAT, 672x1x3x3]
%onnx::Conv_756[FLOAT, 112x672x1x1]
%onnx::Conv_759[FLOAT, 336x112x1x1]
%onnx::Conv_760[FLOAT, 336]
%onnx::Conv_762[FLOAT, 336x1x5x5]
%onnx::Conv_765[FLOAT, 112x336x1x1]
%onnx::Conv_768[FLOAT, 112x112x1x1]
%onnx::Conv_771[FLOAT, 112x1x5x5]
%onnx::Conv_774[FLOAT, 112x112x1x1]
%onnx::Conv_777[FLOAT, 112x112x1x1]
%onnx::Conv_780[FLOAT, 112x1x3x3]
%onnx::Conv_783[FLOAT, 184x112x1x1]
%onnx::Conv_784[FLOAT, 184]
%onnx::Conv_786[FLOAT, 1104x184x1x1]
%onnx::Conv_787[FLOAT, 1104]
%onnx::Conv_789[FLOAT, 1104x1x3x3]
%onnx::Conv_792[FLOAT, 184x1104x1x1]
%onnx::Conv_795[FLOAT, 552x184x1x1]
%onnx::Conv_796[FLOAT, 552]
%onnx::Conv_798[FLOAT, 552x1x3x3]
%onnx::Conv_801[FLOAT, 184x552x1x1]
%onnx::Conv_804[FLOAT, 552x184x1x1]
%onnx::Conv_807[FLOAT, 552x1x3x3]
%onnx::Conv_810[FLOAT, 184x552x1x1]
%onnx::Conv_813[FLOAT, 184x92x1x1]
%onnx::Conv_816[FLOAT, 184x1x3x3]
%onnx::Conv_819[FLOAT, 352x92x1x1]
%onnx::Conv_820[FLOAT, 352]
%onnx::Conv_822[FLOAT, 1504x352x1x1]
%onnx::Conv_823[FLOAT, 1504]
) {
%onnx::Conv_817 = Identity(%onnx::Conv_784)
%onnx::Conv_814 = Identity(%onnx::Conv_784)
%onnx::Conv_811 = Identity(%onnx::Conv_784)
%onnx::Conv_808 = Identity(%onnx::Conv_796)
%onnx::Conv_805 = Identity(%onnx::Conv_796)
%onnx::Conv_802 = Identity(%onnx::Conv_784)
%onnx::Conv_799 = Identity(%onnx::Conv_796)
%onnx::Conv_793 = Identity(%onnx::Conv_784)
%onnx::Conv_790 = Identity(%onnx::Conv_787)
%onnx::Conv_781 = Identity(%onnx::Conv_748)
%onnx::Conv_778 = Identity(%onnx::Conv_748)
%onnx::Conv_775 = Identity(%onnx::Conv_748)
%onnx::Conv_772 = Identity(%onnx::Conv_748)
%onnx::Conv_769 = Identity(%onnx::Conv_748)
%onnx::Conv_766 = Identity(%onnx::Conv_748)
%onnx::Conv_763 = Identity(%onnx::Conv_760)
%onnx::Conv_757 = Identity(%onnx::Conv_748)
%onnx::Conv_754 = Identity(%onnx::Conv_751)
%onnx::Conv_745 = Identity(%onnx::Conv_712)
%onnx::Conv_742 = Identity(%onnx::Conv_712)
%onnx::Conv_739 = Identity(%onnx::Conv_712)
%onnx::Conv_736 = Identity(%onnx::Conv_688)
%onnx::Conv_733 = Identity(%onnx::Conv_688)
%onnx::Conv_730 = Identity(%onnx::Conv_712)
%onnx::Conv_727 = Identity(%onnx::Conv_724)
%onnx::Conv_721 = Identity(%onnx::Conv_712)
%onnx::Conv_718 = Identity(%onnx::Conv_712)
%onnx::Conv_715 = Identity(%onnx::Conv_712)
%onnx::Conv_709 = Identity(%onnx::Conv_688)
%onnx::Conv_706 = Identity(%onnx::Conv_688)
%onnx::Conv_703 = Identity(%onnx::Conv_676)
%onnx::Conv_700 = Identity(%onnx::Conv_643)
%onnx::Conv_697 = Identity(%onnx::Conv_643)
%onnx::Conv_694 = Identity(%onnx::Conv_676)
%onnx::Conv_691 = Identity(%onnx::Conv_688)
%onnx::Conv_685 = Identity(%onnx::Conv_676)
%onnx::Conv_682 = Identity(%onnx::Conv_676)
%onnx::Conv_679 = Identity(%onnx::Conv_676)
%onnx::Conv_673 = Identity(%onnx::Conv_652)
%onnx::Conv_670 = Identity(%onnx::Conv_652)
%onnx::Conv_667 = Identity(%onnx::Conv_649)
%onnx::Conv_664 = Identity(%onnx::Conv_649)
%onnx::Conv_661 = Identity(%onnx::Conv_649)
%onnx::Conv_658 = Identity(%onnx::Conv_649)
%onnx::Conv_655 = Identity(%onnx::Conv_652)
%onnx::Conv_646 = Identity(%onnx::Conv_643)
%onnx::Conv_640 = Identity(%onnx::Conv_631)
%onnx::Conv_637 = Identity(%onnx::Conv_631)
%onnx::Conv_634 = Identity(%onnx::Conv_631)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_630, %onnx::Conv_631)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%628 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %628
}
|
val_accuracy
| 0
| 81,610,368
| 2,074,772
|
{'zcp_synflow': 82.98665932405036, 'zcp_zen': 74.5756607055664, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.17653712630271912, 'zcp_flops': 81610368.0, 'zcp_grad_norm': 27.450136184692383, 'zcp_grasp': 0.1151580810546875, 'zcp_jacov': -16.06051773247095, 'zcp_l2_norm': 708.357666015625, 'zcp_nwot': 216.19588939292387, 'zcp_params': 2074772.0, 'zcp_plain': -0.0005511296913027763, 'zcp_snip': 56.32129669189453, 'lat_1080ti_1': 0.5806200918010798, 'lat_1080ti_32': 0.5633284993424224, 'lat_1080ti_64': 0.5135210082170059, 'lat_2080ti_1': 0.7014146747883179, 'lat_2080ti_32': 0.6219106683494259, 'lat_2080ti_64': 0.5582987209205502, 'lat_essential_ph_1': 0.24528301886792453, 'lat_eyeriss': 0.5830047860928244, 'lat_fpga': 0.6503131833879217, 'lat_gold_6226': 0.4674962843591446, 'lat_gold_6240': 0.6036395554692581, 'lat_pixel2': 0.41304347826086957, 'lat_pixel3': 0.50799245130642, 'lat_raspi4': 0.5274538962636552, 'lat_samsung_a50': 0.2736842105263158, 'lat_samsung_s7': 0.5196850393700787, 'lat_silver_4114': 0.5899082458824654, 'lat_silver_4210r': 0.6294648451985361, 'lat_titan_rtx_1': 0.6285301804853893, 'lat_titan_rtx_32': 0.57385265021544, 'lat_titan_rtx_64': 0.5645263905518388, 'lat_titanx_1': 0.33908324241822285, 'lat_titanx_32': 0.588854038621998, 'lat_titanx_64': 0.500556461144324, 'lat_titanxp_1': 0.6038780578535407, 'lat_titanxp_32': 0.6003996211411788, 'lat_titanxp_64': 0.5208097311628134}
| |
FBNet_3605
|
FBNet
|
3605
|
3605
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_607[FLOAT, 16x3x3x3]
%onnx::Conv_608[FLOAT, 16]
%onnx::Conv_610[FLOAT, 48x16x1x1]
%onnx::Conv_611[FLOAT, 48]
%onnx::Conv_613[FLOAT, 48x1x3x3]
%onnx::Conv_616[FLOAT, 16x48x1x1]
%onnx::Conv_619[FLOAT, 48x16x1x1]
%onnx::Conv_622[FLOAT, 48x1x3x3]
%onnx::Conv_625[FLOAT, 24x48x1x1]
%onnx::Conv_626[FLOAT, 24]
%onnx::Conv_628[FLOAT, 24x24x1x1]
%onnx::Conv_631[FLOAT, 24x1x3x3]
%onnx::Conv_634[FLOAT, 24x24x1x1]
%onnx::Conv_637[FLOAT, 24x12x1x1]
%onnx::Conv_640[FLOAT, 24x1x5x5]
%onnx::Conv_643[FLOAT, 24x12x1x1]
%onnx::Conv_646[FLOAT, 24x24x1x1]
%onnx::Conv_649[FLOAT, 24x1x3x3]
%onnx::Conv_652[FLOAT, 32x24x1x1]
%onnx::Conv_653[FLOAT, 32]
%onnx::Conv_655[FLOAT, 192x32x1x1]
%onnx::Conv_656[FLOAT, 192]
%onnx::Conv_658[FLOAT, 192x1x3x3]
%onnx::Conv_661[FLOAT, 32x192x1x1]
%onnx::Conv_664[FLOAT, 192x32x1x1]
%onnx::Conv_667[FLOAT, 192x1x3x3]
%onnx::Conv_670[FLOAT, 32x192x1x1]
%onnx::Conv_673[FLOAT, 32x16x1x1]
%onnx::Conv_676[FLOAT, 32x1x3x3]
%onnx::Conv_679[FLOAT, 32x16x1x1]
%onnx::Conv_682[FLOAT, 32x32x1x1]
%onnx::Conv_685[FLOAT, 32x1x5x5]
%onnx::Conv_688[FLOAT, 64x32x1x1]
%onnx::Conv_689[FLOAT, 64]
%onnx::Conv_691[FLOAT, 64x64x1x1]
%onnx::Conv_694[FLOAT, 64x1x5x5]
%onnx::Conv_697[FLOAT, 64x64x1x1]
%onnx::Conv_700[FLOAT, 64x64x1x1]
%onnx::Conv_703[FLOAT, 64x1x3x3]
%onnx::Conv_706[FLOAT, 64x64x1x1]
%onnx::Conv_709[FLOAT, 384x64x1x1]
%onnx::Conv_710[FLOAT, 384]
%onnx::Conv_712[FLOAT, 384x1x5x5]
%onnx::Conv_715[FLOAT, 64x384x1x1]
%onnx::Conv_718[FLOAT, 112x64x1x1]
%onnx::Conv_719[FLOAT, 112]
%onnx::Conv_721[FLOAT, 672x112x1x1]
%onnx::Conv_722[FLOAT, 672]
%onnx::Conv_724[FLOAT, 672x1x5x5]
%onnx::Conv_727[FLOAT, 112x672x1x1]
%onnx::Conv_730[FLOAT, 112x56x1x1]
%onnx::Conv_733[FLOAT, 112x1x3x3]
%onnx::Conv_736[FLOAT, 112x56x1x1]
%onnx::Conv_739[FLOAT, 112x112x1x1]
%onnx::Conv_742[FLOAT, 112x1x5x5]
%onnx::Conv_745[FLOAT, 112x112x1x1]
%onnx::Conv_748[FLOAT, 336x112x1x1]
%onnx::Conv_749[FLOAT, 336]
%onnx::Conv_751[FLOAT, 336x1x3x3]
%onnx::Conv_754[FLOAT, 184x336x1x1]
%onnx::Conv_755[FLOAT, 184]
%onnx::Conv_757[FLOAT, 184x92x1x1]
%onnx::Conv_760[FLOAT, 184x1x5x5]
%onnx::Conv_763[FLOAT, 184x92x1x1]
%onnx::Conv_766[FLOAT, 552x184x1x1]
%onnx::Conv_767[FLOAT, 552]
%onnx::Conv_769[FLOAT, 552x1x5x5]
%onnx::Conv_772[FLOAT, 184x552x1x1]
%onnx::Conv_775[FLOAT, 352x184x1x1]
%onnx::Conv_776[FLOAT, 352]
%onnx::Conv_778[FLOAT, 1504x352x1x1]
%onnx::Conv_779[FLOAT, 1504]
) {
%onnx::Conv_773 = Identity(%onnx::Conv_755)
%onnx::Conv_770 = Identity(%onnx::Conv_767)
%onnx::Conv_764 = Identity(%onnx::Conv_755)
%onnx::Conv_761 = Identity(%onnx::Conv_755)
%onnx::Conv_758 = Identity(%onnx::Conv_755)
%onnx::Conv_752 = Identity(%onnx::Conv_749)
%onnx::Conv_746 = Identity(%onnx::Conv_719)
%onnx::Conv_743 = Identity(%onnx::Conv_719)
%onnx::Conv_740 = Identity(%onnx::Conv_719)
%onnx::Conv_737 = Identity(%onnx::Conv_719)
%onnx::Conv_734 = Identity(%onnx::Conv_719)
%onnx::Conv_731 = Identity(%onnx::Conv_719)
%onnx::Conv_728 = Identity(%onnx::Conv_719)
%onnx::Conv_725 = Identity(%onnx::Conv_722)
%onnx::Conv_716 = Identity(%onnx::Conv_689)
%onnx::Conv_713 = Identity(%onnx::Conv_710)
%onnx::Conv_707 = Identity(%onnx::Conv_689)
%onnx::Conv_704 = Identity(%onnx::Conv_689)
%onnx::Conv_701 = Identity(%onnx::Conv_689)
%onnx::Conv_698 = Identity(%onnx::Conv_689)
%onnx::Conv_695 = Identity(%onnx::Conv_689)
%onnx::Conv_692 = Identity(%onnx::Conv_689)
%onnx::Conv_686 = Identity(%onnx::Conv_653)
%onnx::Conv_683 = Identity(%onnx::Conv_653)
%onnx::Conv_680 = Identity(%onnx::Conv_653)
%onnx::Conv_677 = Identity(%onnx::Conv_653)
%onnx::Conv_674 = Identity(%onnx::Conv_653)
%onnx::Conv_671 = Identity(%onnx::Conv_653)
%onnx::Conv_668 = Identity(%onnx::Conv_656)
%onnx::Conv_665 = Identity(%onnx::Conv_656)
%onnx::Conv_662 = Identity(%onnx::Conv_653)
%onnx::Conv_659 = Identity(%onnx::Conv_656)
%onnx::Conv_650 = Identity(%onnx::Conv_626)
%onnx::Conv_647 = Identity(%onnx::Conv_626)
%onnx::Conv_644 = Identity(%onnx::Conv_626)
%onnx::Conv_641 = Identity(%onnx::Conv_626)
%onnx::Conv_638 = Identity(%onnx::Conv_626)
%onnx::Conv_635 = Identity(%onnx::Conv_626)
%onnx::Conv_632 = Identity(%onnx::Conv_626)
%onnx::Conv_629 = Identity(%onnx::Conv_626)
%onnx::Conv_623 = Identity(%onnx::Conv_611)
%onnx::Conv_620 = Identity(%onnx::Conv_611)
%onnx::Conv_617 = Identity(%onnx::Conv_608)
%onnx::Conv_614 = Identity(%onnx::Conv_611)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_607, %onnx::Conv_608)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_610, %onnx::Conv_611)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_613, %onnx::Conv_614)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_616, %onnx::Conv_617)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_619, %onnx::Conv_620)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_622, %onnx::Conv_623)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_625, %onnx::Conv_626)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_628, %onnx::Conv_629)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_631, %onnx::Conv_632)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_634, %onnx::Conv_635)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_637, %onnx::Conv_638)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_640, %onnx::Conv_641)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_643, %onnx::Conv_644)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_646, %onnx::Conv_647)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_649, %onnx::Conv_650)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_652, %onnx::Conv_653)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_655, %onnx::Conv_656)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_658, %onnx::Conv_659)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_output_0 = Reshape(%/cells.8/nl/Relu_output_0, %/cells.8/shuffle/Constant_output_0)
%/cells.8/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.8/shuffle/Reshape_output_0)
%/cells.8/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.8/shuffle/Reshape_1_output_0 = Reshape(%/cells.8/shuffle/Transpose_output_0, %/cells.8/shuffle/Constant_1_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/shuffle/Reshape_1_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.13/relu/Relu_output_0 = Relu(%/cells.13/conv/Conv_output_0)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/relu/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/relu/Relu_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_output_0 = Reshape(%/cells.15/nl/Relu_output_0, %/cells.15/shuffle/Constant_output_0)
%/cells.15/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.15/shuffle/Reshape_output_0)
%/cells.15/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.15/shuffle/Reshape_1_output_0 = Reshape(%/cells.15/shuffle/Transpose_output_0, %/cells.15/shuffle/Constant_1_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/shuffle/Reshape_1_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.21/relu/Relu_output_0 = Relu(%/cells.21/conv/Conv_output_0)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/relu/Relu_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%605 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %605
}
|
val_accuracy
| 0
| 54,678,144
| 1,456,940
|
{'zcp_synflow': 73.48524032827244, 'zcp_zen': 61.016029357910156, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.060714203864336014, 'zcp_flops': 54678144.0, 'zcp_grad_norm': 19.15534019470215, 'zcp_grasp': -0.041396141052246094, 'zcp_jacov': -16.06318188039884, 'zcp_l2_norm': 543.4249267578125, 'zcp_nwot': 208.49644810664591, 'zcp_params': 1456940.0, 'zcp_plain': 0.002870428841561079, 'zcp_snip': 29.282150268554688, 'lat_1080ti_1': 0.46603741299999124, 'lat_1080ti_32': 0.31474219700645495, 'lat_1080ti_64': 0.2586013791137018, 'lat_2080ti_1': 0.48287285222116383, 'lat_2080ti_32': 0.3551383299430393, 'lat_2080ti_64': 0.2694958865901798, 'lat_essential_ph_1': 0.09433962264150944, 'lat_eyeriss': 0.2617864412374157, 'lat_fpga': 0.2745722795594652, 'lat_gold_6226': 0.23068917356047142, 'lat_gold_6240': 0.33193605831517825, 'lat_pixel2': 0.17391304347826086, 'lat_pixel3': 0.2547028523942694, 'lat_raspi4': 0.22665807700561624, 'lat_samsung_a50': 0.10526315789473684, 'lat_samsung_s7': 0.14173228346456693, 'lat_silver_4114': 0.37140004688948863, 'lat_silver_4210r': 0.3842739904596354, 'lat_titan_rtx_1': 0.4533313425583886, 'lat_titan_rtx_32': 0.3531146226101284, 'lat_titan_rtx_64': 0.2743466177393521, 'lat_titanx_1': 0.2344000324165826, 'lat_titanx_32': 0.26393588623342285, 'lat_titanx_64': 0.2571382687884053, 'lat_titanxp_1': 0.43453108749798547, 'lat_titanxp_32': 0.3127861011364877, 'lat_titanxp_64': 0.2614624902688626}
| |
FBNet_832
|
FBNet
|
832
|
832
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_588[FLOAT, 16x3x3x3]
%onnx::Conv_589[FLOAT, 16]
%onnx::Conv_591[FLOAT, 96x16x1x1]
%onnx::Conv_592[FLOAT, 96]
%onnx::Conv_594[FLOAT, 96x1x3x3]
%onnx::Conv_597[FLOAT, 16x96x1x1]
%onnx::Conv_600[FLOAT, 16x8x1x1]
%onnx::Conv_603[FLOAT, 16x1x3x3]
%onnx::Conv_606[FLOAT, 24x8x1x1]
%onnx::Conv_607[FLOAT, 24]
%onnx::Conv_609[FLOAT, 24x12x1x1]
%onnx::Conv_612[FLOAT, 24x1x3x3]
%onnx::Conv_615[FLOAT, 24x12x1x1]
%onnx::Conv_618[FLOAT, 72x24x1x1]
%onnx::Conv_619[FLOAT, 72]
%onnx::Conv_621[FLOAT, 72x1x5x5]
%onnx::Conv_624[FLOAT, 24x72x1x1]
%onnx::Conv_627[FLOAT, 24x24x1x1]
%onnx::Conv_630[FLOAT, 24x1x3x3]
%onnx::Conv_633[FLOAT, 24x24x1x1]
%onnx::Conv_636[FLOAT, 72x24x1x1]
%onnx::Conv_639[FLOAT, 72x1x5x5]
%onnx::Conv_642[FLOAT, 32x72x1x1]
%onnx::Conv_643[FLOAT, 32]
%onnx::Conv_645[FLOAT, 192x32x1x1]
%onnx::Conv_646[FLOAT, 192]
%onnx::Conv_648[FLOAT, 192x1x5x5]
%onnx::Conv_651[FLOAT, 32x192x1x1]
%onnx::Conv_654[FLOAT, 96x32x1x1]
%onnx::Conv_657[FLOAT, 96x1x3x3]
%onnx::Conv_660[FLOAT, 32x96x1x1]
%onnx::Conv_663[FLOAT, 96x32x1x1]
%onnx::Conv_666[FLOAT, 96x1x5x5]
%onnx::Conv_669[FLOAT, 32x96x1x1]
%onnx::Conv_672[FLOAT, 192x32x1x1]
%onnx::Conv_675[FLOAT, 192x1x3x3]
%onnx::Conv_678[FLOAT, 64x192x1x1]
%onnx::Conv_679[FLOAT, 64]
%onnx::Conv_681[FLOAT, 384x64x1x1]
%onnx::Conv_682[FLOAT, 384]
%onnx::Conv_684[FLOAT, 384x1x5x5]
%onnx::Conv_687[FLOAT, 64x384x1x1]
%onnx::Conv_690[FLOAT, 384x64x1x1]
%onnx::Conv_693[FLOAT, 384x1x3x3]
%onnx::Conv_696[FLOAT, 64x384x1x1]
%onnx::Conv_699[FLOAT, 64x32x1x1]
%onnx::Conv_702[FLOAT, 64x1x3x3]
%onnx::Conv_705[FLOAT, 64x32x1x1]
%onnx::Conv_708[FLOAT, 112x64x1x1]
%onnx::Conv_709[FLOAT, 112]
%onnx::Conv_711[FLOAT, 672x112x1x1]
%onnx::Conv_712[FLOAT, 672]
%onnx::Conv_714[FLOAT, 672x1x5x5]
%onnx::Conv_717[FLOAT, 112x672x1x1]
%onnx::Conv_720[FLOAT, 112x112x1x1]
%onnx::Conv_723[FLOAT, 112x1x5x5]
%onnx::Conv_726[FLOAT, 184x112x1x1]
%onnx::Conv_727[FLOAT, 184]
%onnx::Conv_729[FLOAT, 552x184x1x1]
%onnx::Conv_730[FLOAT, 552]
%onnx::Conv_732[FLOAT, 552x1x5x5]
%onnx::Conv_735[FLOAT, 184x552x1x1]
%onnx::Conv_738[FLOAT, 1104x184x1x1]
%onnx::Conv_739[FLOAT, 1104]
%onnx::Conv_741[FLOAT, 1104x1x3x3]
%onnx::Conv_744[FLOAT, 184x1104x1x1]
%onnx::Conv_747[FLOAT, 1104x184x1x1]
%onnx::Conv_750[FLOAT, 1104x1x3x3]
%onnx::Conv_753[FLOAT, 184x1104x1x1]
%onnx::Conv_756[FLOAT, 352x184x1x1]
%onnx::Conv_757[FLOAT, 352]
%onnx::Conv_759[FLOAT, 1504x352x1x1]
%onnx::Conv_760[FLOAT, 1504]
) {
%onnx::Conv_754 = Identity(%onnx::Conv_727)
%onnx::Conv_751 = Identity(%onnx::Conv_739)
%onnx::Conv_748 = Identity(%onnx::Conv_739)
%onnx::Conv_745 = Identity(%onnx::Conv_727)
%onnx::Conv_742 = Identity(%onnx::Conv_739)
%onnx::Conv_736 = Identity(%onnx::Conv_727)
%onnx::Conv_733 = Identity(%onnx::Conv_730)
%onnx::Conv_724 = Identity(%onnx::Conv_709)
%onnx::Conv_721 = Identity(%onnx::Conv_709)
%onnx::Conv_718 = Identity(%onnx::Conv_709)
%onnx::Conv_715 = Identity(%onnx::Conv_712)
%onnx::Conv_706 = Identity(%onnx::Conv_679)
%onnx::Conv_703 = Identity(%onnx::Conv_679)
%onnx::Conv_700 = Identity(%onnx::Conv_679)
%onnx::Conv_697 = Identity(%onnx::Conv_679)
%onnx::Conv_694 = Identity(%onnx::Conv_682)
%onnx::Conv_691 = Identity(%onnx::Conv_682)
%onnx::Conv_688 = Identity(%onnx::Conv_679)
%onnx::Conv_685 = Identity(%onnx::Conv_682)
%onnx::Conv_676 = Identity(%onnx::Conv_646)
%onnx::Conv_673 = Identity(%onnx::Conv_646)
%onnx::Conv_670 = Identity(%onnx::Conv_643)
%onnx::Conv_667 = Identity(%onnx::Conv_592)
%onnx::Conv_664 = Identity(%onnx::Conv_592)
%onnx::Conv_661 = Identity(%onnx::Conv_643)
%onnx::Conv_658 = Identity(%onnx::Conv_592)
%onnx::Conv_655 = Identity(%onnx::Conv_592)
%onnx::Conv_652 = Identity(%onnx::Conv_643)
%onnx::Conv_649 = Identity(%onnx::Conv_646)
%onnx::Conv_640 = Identity(%onnx::Conv_619)
%onnx::Conv_637 = Identity(%onnx::Conv_619)
%onnx::Conv_634 = Identity(%onnx::Conv_607)
%onnx::Conv_631 = Identity(%onnx::Conv_607)
%onnx::Conv_628 = Identity(%onnx::Conv_607)
%onnx::Conv_625 = Identity(%onnx::Conv_607)
%onnx::Conv_622 = Identity(%onnx::Conv_619)
%onnx::Conv_616 = Identity(%onnx::Conv_607)
%onnx::Conv_613 = Identity(%onnx::Conv_607)
%onnx::Conv_610 = Identity(%onnx::Conv_607)
%onnx::Conv_604 = Identity(%onnx::Conv_589)
%onnx::Conv_601 = Identity(%onnx::Conv_589)
%onnx::Conv_598 = Identity(%onnx::Conv_589)
%onnx::Conv_595 = Identity(%onnx::Conv_592)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_588, %onnx::Conv_589)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_591, %onnx::Conv_592)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_594, %onnx::Conv_595)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_597, %onnx::Conv_598)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_600, %onnx::Conv_601)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_603, %onnx::Conv_604)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_606, %onnx::Conv_607)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_609, %onnx::Conv_610)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_612, %onnx::Conv_613)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_615, %onnx::Conv_616)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_618, %onnx::Conv_619)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_621, %onnx::Conv_622)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_624, %onnx::Conv_625)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_627, %onnx::Conv_628)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_630, %onnx::Conv_631)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.13/relu/Relu_output_0 = Relu(%/cells.13/conv/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/relu/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.13/relu/Relu_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.21/relu/Relu_output_0 = Relu(%/cells.21/conv/Conv_output_0)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/relu/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%586 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %586
}
|
val_accuracy
| 0
| 75,694,208
| 2,219,068
|
{'zcp_synflow': 74.43211582487888, 'zcp_zen': 65.0946044921875, 'zcp_epe_nas': 0.00015999920000638146, 'zcp_fisher': 0.06287675350904465, 'zcp_flops': 75694208.0, 'zcp_grad_norm': 21.275543212890625, 'zcp_grasp': -0.020660400390625, 'zcp_jacov': -16.061922502495797, 'zcp_l2_norm': 633.3408203125, 'zcp_nwot': 214.1274071688382, 'zcp_params': 2219068.0, 'zcp_plain': -0.001514284871518612, 'zcp_snip': 34.45809555053711, 'lat_1080ti_1': 0.3996030877934451, 'lat_1080ti_32': 0.39200119341027256, 'lat_1080ti_64': 0.37559850822483615, 'lat_2080ti_1': 0.4366015112873124, 'lat_2080ti_32': 0.3694096402665864, 'lat_2080ti_64': 0.3596758345728778, 'lat_essential_ph_1': 0.39622641509433965, 'lat_eyeriss': 0.5680780444868246, 'lat_fpga': 0.5399024332320721, 'lat_gold_6226': 0.49890417363309264, 'lat_gold_6240': 0.5836235575805307, 'lat_pixel2': 0.41304347826086957, 'lat_pixel3': 0.5308377694639888, 'lat_raspi4': 0.5377290725436202, 'lat_samsung_a50': 0.23157894736842105, 'lat_samsung_s7': 0.2125984251968504, 'lat_silver_4114': 0.6645552662532054, 'lat_silver_4210r': 0.5700021991659917, 'lat_titan_rtx_1': 0.4254705163637974, 'lat_titan_rtx_32': 0.3549743929350933, 'lat_titan_rtx_64': 0.3671488114298684, 'lat_titanx_1': 0.2337008272230176, 'lat_titanx_32': 0.35789149499523754, 'lat_titanx_64': 0.3970111784726104, 'lat_titanxp_1': 0.42861545552493646, 'lat_titanxp_32': 0.36288101727894284, 'lat_titanxp_64': 0.3820427741233394}
| |
FBNet_2310
|
FBNet
|
2310
|
2310
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_586[FLOAT, 16x3x3x3]
%onnx::Conv_587[FLOAT, 16]
%onnx::Conv_589[FLOAT, 16x8x1x1]
%onnx::Conv_592[FLOAT, 16x1x5x5]
%onnx::Conv_595[FLOAT, 16x8x1x1]
%onnx::Conv_598[FLOAT, 16x8x1x1]
%onnx::Conv_601[FLOAT, 16x1x5x5]
%onnx::Conv_604[FLOAT, 24x8x1x1]
%onnx::Conv_605[FLOAT, 24]
%onnx::Conv_607[FLOAT, 24x24x1x1]
%onnx::Conv_610[FLOAT, 24x1x3x3]
%onnx::Conv_613[FLOAT, 24x24x1x1]
%onnx::Conv_616[FLOAT, 72x24x1x1]
%onnx::Conv_617[FLOAT, 72]
%onnx::Conv_619[FLOAT, 72x1x5x5]
%onnx::Conv_622[FLOAT, 24x72x1x1]
%onnx::Conv_625[FLOAT, 72x24x1x1]
%onnx::Conv_628[FLOAT, 72x1x3x3]
%onnx::Conv_631[FLOAT, 24x72x1x1]
%onnx::Conv_634[FLOAT, 144x24x1x1]
%onnx::Conv_635[FLOAT, 144]
%onnx::Conv_637[FLOAT, 144x1x5x5]
%onnx::Conv_640[FLOAT, 32x144x1x1]
%onnx::Conv_641[FLOAT, 32]
%onnx::Conv_643[FLOAT, 192x32x1x1]
%onnx::Conv_644[FLOAT, 192]
%onnx::Conv_646[FLOAT, 192x1x3x3]
%onnx::Conv_649[FLOAT, 32x192x1x1]
%onnx::Conv_652[FLOAT, 96x32x1x1]
%onnx::Conv_653[FLOAT, 96]
%onnx::Conv_655[FLOAT, 96x1x3x3]
%onnx::Conv_658[FLOAT, 32x96x1x1]
%onnx::Conv_661[FLOAT, 32x16x1x1]
%onnx::Conv_664[FLOAT, 32x1x5x5]
%onnx::Conv_667[FLOAT, 64x16x1x1]
%onnx::Conv_668[FLOAT, 64]
%onnx::Conv_670[FLOAT, 64x32x1x1]
%onnx::Conv_673[FLOAT, 64x1x3x3]
%onnx::Conv_676[FLOAT, 64x32x1x1]
%onnx::Conv_679[FLOAT, 384x64x1x1]
%onnx::Conv_680[FLOAT, 384]
%onnx::Conv_682[FLOAT, 384x1x3x3]
%onnx::Conv_685[FLOAT, 64x384x1x1]
%onnx::Conv_688[FLOAT, 384x64x1x1]
%onnx::Conv_691[FLOAT, 384x1x5x5]
%onnx::Conv_694[FLOAT, 64x384x1x1]
%onnx::Conv_697[FLOAT, 64x64x1x1]
%onnx::Conv_700[FLOAT, 64x1x3x3]
%onnx::Conv_703[FLOAT, 112x64x1x1]
%onnx::Conv_704[FLOAT, 112]
%onnx::Conv_706[FLOAT, 112x112x1x1]
%onnx::Conv_709[FLOAT, 112x1x3x3]
%onnx::Conv_712[FLOAT, 112x112x1x1]
%onnx::Conv_715[FLOAT, 112x112x1x1]
%onnx::Conv_718[FLOAT, 112x1x3x3]
%onnx::Conv_721[FLOAT, 184x112x1x1]
%onnx::Conv_722[FLOAT, 184]
%onnx::Conv_724[FLOAT, 552x184x1x1]
%onnx::Conv_725[FLOAT, 552]
%onnx::Conv_727[FLOAT, 552x1x3x3]
%onnx::Conv_730[FLOAT, 184x552x1x1]
%onnx::Conv_733[FLOAT, 552x184x1x1]
%onnx::Conv_736[FLOAT, 552x1x5x5]
%onnx::Conv_739[FLOAT, 184x552x1x1]
%onnx::Conv_742[FLOAT, 184x184x1x1]
%onnx::Conv_745[FLOAT, 184x1x3x3]
%onnx::Conv_748[FLOAT, 352x184x1x1]
%onnx::Conv_749[FLOAT, 352]
%onnx::Conv_751[FLOAT, 1504x352x1x1]
%onnx::Conv_752[FLOAT, 1504]
) {
%onnx::Conv_746 = Identity(%onnx::Conv_722)
%onnx::Conv_743 = Identity(%onnx::Conv_722)
%onnx::Conv_740 = Identity(%onnx::Conv_722)
%onnx::Conv_737 = Identity(%onnx::Conv_725)
%onnx::Conv_734 = Identity(%onnx::Conv_725)
%onnx::Conv_731 = Identity(%onnx::Conv_722)
%onnx::Conv_728 = Identity(%onnx::Conv_725)
%onnx::Conv_719 = Identity(%onnx::Conv_704)
%onnx::Conv_716 = Identity(%onnx::Conv_704)
%onnx::Conv_713 = Identity(%onnx::Conv_704)
%onnx::Conv_710 = Identity(%onnx::Conv_704)
%onnx::Conv_707 = Identity(%onnx::Conv_704)
%onnx::Conv_701 = Identity(%onnx::Conv_668)
%onnx::Conv_698 = Identity(%onnx::Conv_668)
%onnx::Conv_695 = Identity(%onnx::Conv_668)
%onnx::Conv_692 = Identity(%onnx::Conv_680)
%onnx::Conv_689 = Identity(%onnx::Conv_680)
%onnx::Conv_686 = Identity(%onnx::Conv_668)
%onnx::Conv_683 = Identity(%onnx::Conv_680)
%onnx::Conv_677 = Identity(%onnx::Conv_668)
%onnx::Conv_674 = Identity(%onnx::Conv_668)
%onnx::Conv_671 = Identity(%onnx::Conv_668)
%onnx::Conv_665 = Identity(%onnx::Conv_641)
%onnx::Conv_662 = Identity(%onnx::Conv_641)
%onnx::Conv_659 = Identity(%onnx::Conv_641)
%onnx::Conv_656 = Identity(%onnx::Conv_653)
%onnx::Conv_650 = Identity(%onnx::Conv_641)
%onnx::Conv_647 = Identity(%onnx::Conv_644)
%onnx::Conv_638 = Identity(%onnx::Conv_635)
%onnx::Conv_632 = Identity(%onnx::Conv_605)
%onnx::Conv_629 = Identity(%onnx::Conv_617)
%onnx::Conv_626 = Identity(%onnx::Conv_617)
%onnx::Conv_623 = Identity(%onnx::Conv_605)
%onnx::Conv_620 = Identity(%onnx::Conv_617)
%onnx::Conv_614 = Identity(%onnx::Conv_605)
%onnx::Conv_611 = Identity(%onnx::Conv_605)
%onnx::Conv_608 = Identity(%onnx::Conv_605)
%onnx::Conv_602 = Identity(%onnx::Conv_587)
%onnx::Conv_599 = Identity(%onnx::Conv_587)
%onnx::Conv_596 = Identity(%onnx::Conv_587)
%onnx::Conv_593 = Identity(%onnx::Conv_587)
%onnx::Conv_590 = Identity(%onnx::Conv_587)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_586, %onnx::Conv_587)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_589, %onnx::Conv_590)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_592, %onnx::Conv_593)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_595, %onnx::Conv_596)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_598, %onnx::Conv_599)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_601, %onnx::Conv_602)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_604, %onnx::Conv_605)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_607, %onnx::Conv_608)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_610, %onnx::Conv_611)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_613, %onnx::Conv_614)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_616, %onnx::Conv_617)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_619, %onnx::Conv_620)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_622, %onnx::Conv_623)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_625, %onnx::Conv_626)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_628, %onnx::Conv_629)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_631, %onnx::Conv_632)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_634, %onnx::Conv_635)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_637, %onnx::Conv_638)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_640, %onnx::Conv_641)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_643, %onnx::Conv_644)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_646, %onnx::Conv_647)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_649, %onnx::Conv_650)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_652, %onnx::Conv_653)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_655, %onnx::Conv_656)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_658, %onnx::Conv_659)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_661, %onnx::Conv_662)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_output_0 = Reshape(%/cells.9/nl/Relu_output_0, %/cells.9/shuffle/Constant_output_0)
%/cells.9/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.9/shuffle/Reshape_output_0)
%/cells.9/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.9/shuffle/Reshape_1_output_0 = Reshape(%/cells.9/shuffle/Transpose_output_0, %/cells.9/shuffle/Constant_1_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/shuffle/Reshape_1_output_0, %onnx::Conv_664, %onnx::Conv_665)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_667, %onnx::Conv_668)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_670, %onnx::Conv_671)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_output_0 = Reshape(%/cells.10/nl/Relu_output_0, %/cells.10/shuffle/Constant_output_0)
%/cells.10/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.10/shuffle/Reshape_output_0)
%/cells.10/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.10/shuffle/Reshape_1_output_0 = Reshape(%/cells.10/shuffle/Transpose_output_0, %/cells.10/shuffle/Constant_1_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/shuffle/Reshape_1_output_0, %onnx::Conv_673, %onnx::Conv_674)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_676, %onnx::Conv_677)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_679, %onnx::Conv_680)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_682, %onnx::Conv_683)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_685, %onnx::Conv_686)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_688, %onnx::Conv_689)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_691, %onnx::Conv_692)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_694, %onnx::Conv_695)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_697, %onnx::Conv_698)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_700, %onnx::Conv_701)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_703, %onnx::Conv_704)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_706, %onnx::Conv_707)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_709, %onnx::Conv_710)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_712, %onnx::Conv_713)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_715, %onnx::Conv_716)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%584 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %584
}
|
val_accuracy
| 0
| 54,915,968
| 1,459,428
|
{'zcp_synflow': 68.1268740179987, 'zcp_zen': 58.78299331665039, 'zcp_epe_nas': 16.42403535828298, 'zcp_fisher': 0.05590430274605751, 'zcp_flops': 54915968.0, 'zcp_grad_norm': 18.769493103027344, 'zcp_grasp': -0.03134346008300781, 'zcp_jacov': -16.066786322829962, 'zcp_l2_norm': 527.4993286132812, 'zcp_nwot': 211.28304069964744, 'zcp_params': 1459428.0, 'zcp_plain': 0.004039539489895105, 'zcp_snip': 29.849082946777344, 'lat_1080ti_1': 0.38467954494670387, 'lat_1080ti_32': 0.33391681644330445, 'lat_1080ti_64': 0.3261917140830497, 'lat_2080ti_1': 0.38969066549316905, 'lat_2080ti_32': 0.33862706992247726, 'lat_2080ti_64': 0.3482071990202379, 'lat_essential_ph_1': 0.20754716981132076, 'lat_eyeriss': 0.3082078741312601, 'lat_fpga': 0.27619509395501635, 'lat_gold_6226': 0.2246059146399063, 'lat_gold_6240': 0.2626277471840641, 'lat_pixel2': 0.17391304347826086, 'lat_pixel3': 0.30264675949354136, 'lat_raspi4': 0.28142588685677106, 'lat_samsung_a50': 0.12631578947368421, 'lat_samsung_s7': 0.10236220472440945, 'lat_silver_4114': 0.2690271463497547, 'lat_silver_4210r': 0.2549309479685939, 'lat_titan_rtx_1': 0.36808923620565703, 'lat_titan_rtx_32': 0.3050678868243458, 'lat_titan_rtx_64': 0.32844637484510686, 'lat_titanx_1': 0.19319786039496753, 'lat_titanx_32': 0.3178794592632878, 'lat_titanx_64': 0.33305929991618805, 'lat_titanxp_1': 0.36963175287359223, 'lat_titanxp_32': 0.32924196317750276, 'lat_titanxp_64': 0.33860191167210046}
| |
FBNet_4114
|
FBNet
|
4114
|
4114
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_660[FLOAT, 16x3x3x3]
%onnx::Conv_661[FLOAT, 16]
%onnx::Conv_663[FLOAT, 16x8x1x1]
%onnx::Conv_666[FLOAT, 16x1x5x5]
%onnx::Conv_669[FLOAT, 16x8x1x1]
%onnx::Conv_672[FLOAT, 96x16x1x1]
%onnx::Conv_673[FLOAT, 96]
%onnx::Conv_675[FLOAT, 96x1x5x5]
%onnx::Conv_678[FLOAT, 24x96x1x1]
%onnx::Conv_679[FLOAT, 24]
%onnx::Conv_681[FLOAT, 24x12x1x1]
%onnx::Conv_684[FLOAT, 24x1x3x3]
%onnx::Conv_687[FLOAT, 24x12x1x1]
%onnx::Conv_690[FLOAT, 24x24x1x1]
%onnx::Conv_693[FLOAT, 24x1x5x5]
%onnx::Conv_696[FLOAT, 24x24x1x1]
%onnx::Conv_699[FLOAT, 24x12x1x1]
%onnx::Conv_702[FLOAT, 24x1x3x3]
%onnx::Conv_705[FLOAT, 32x12x1x1]
%onnx::Conv_706[FLOAT, 32]
%onnx::Conv_708[FLOAT, 96x32x1x1]
%onnx::Conv_711[FLOAT, 96x1x3x3]
%onnx::Conv_714[FLOAT, 32x96x1x1]
%onnx::Conv_717[FLOAT, 32x32x1x1]
%onnx::Conv_720[FLOAT, 32x1x3x3]
%onnx::Conv_723[FLOAT, 32x32x1x1]
%onnx::Conv_726[FLOAT, 96x32x1x1]
%onnx::Conv_729[FLOAT, 96x1x3x3]
%onnx::Conv_732[FLOAT, 64x96x1x1]
%onnx::Conv_733[FLOAT, 64]
%onnx::Conv_735[FLOAT, 384x64x1x1]
%onnx::Conv_736[FLOAT, 384]
%onnx::Conv_738[FLOAT, 384x1x5x5]
%onnx::Conv_741[FLOAT, 64x384x1x1]
%onnx::Conv_744[FLOAT, 64x32x1x1]
%onnx::Conv_747[FLOAT, 64x1x3x3]
%onnx::Conv_750[FLOAT, 64x32x1x1]
%onnx::Conv_753[FLOAT, 64x32x1x1]
%onnx::Conv_756[FLOAT, 64x1x3x3]
%onnx::Conv_759[FLOAT, 64x32x1x1]
%onnx::Conv_762[FLOAT, 112x64x1x1]
%onnx::Conv_763[FLOAT, 112]
%onnx::Conv_765[FLOAT, 112x112x1x1]
%onnx::Conv_768[FLOAT, 112x1x3x3]
%onnx::Conv_771[FLOAT, 112x112x1x1]
%onnx::Conv_774[FLOAT, 336x112x1x1]
%onnx::Conv_775[FLOAT, 336]
%onnx::Conv_777[FLOAT, 336x1x5x5]
%onnx::Conv_780[FLOAT, 112x336x1x1]
%onnx::Conv_783[FLOAT, 336x112x1x1]
%onnx::Conv_786[FLOAT, 336x1x3x3]
%onnx::Conv_789[FLOAT, 112x336x1x1]
%onnx::Conv_792[FLOAT, 112x112x1x1]
%onnx::Conv_795[FLOAT, 112x1x5x5]
%onnx::Conv_798[FLOAT, 184x112x1x1]
%onnx::Conv_799[FLOAT, 184]
%onnx::Conv_801[FLOAT, 1104x184x1x1]
%onnx::Conv_802[FLOAT, 1104]
%onnx::Conv_804[FLOAT, 1104x1x3x3]
%onnx::Conv_807[FLOAT, 184x1104x1x1]
%onnx::Conv_810[FLOAT, 1104x184x1x1]
%onnx::Conv_813[FLOAT, 1104x1x3x3]
%onnx::Conv_816[FLOAT, 184x1104x1x1]
%onnx::Conv_819[FLOAT, 1104x184x1x1]
%onnx::Conv_822[FLOAT, 1104x1x3x3]
%onnx::Conv_825[FLOAT, 184x1104x1x1]
%onnx::Conv_828[FLOAT, 184x92x1x1]
%onnx::Conv_831[FLOAT, 184x1x3x3]
%onnx::Conv_834[FLOAT, 352x92x1x1]
%onnx::Conv_835[FLOAT, 352]
%onnx::Conv_837[FLOAT, 1504x352x1x1]
%onnx::Conv_838[FLOAT, 1504]
) {
%onnx::Conv_832 = Identity(%onnx::Conv_799)
%onnx::Conv_829 = Identity(%onnx::Conv_799)
%onnx::Conv_826 = Identity(%onnx::Conv_799)
%onnx::Conv_823 = Identity(%onnx::Conv_802)
%onnx::Conv_820 = Identity(%onnx::Conv_802)
%onnx::Conv_817 = Identity(%onnx::Conv_799)
%onnx::Conv_814 = Identity(%onnx::Conv_802)
%onnx::Conv_811 = Identity(%onnx::Conv_802)
%onnx::Conv_808 = Identity(%onnx::Conv_799)
%onnx::Conv_805 = Identity(%onnx::Conv_802)
%onnx::Conv_796 = Identity(%onnx::Conv_763)
%onnx::Conv_793 = Identity(%onnx::Conv_763)
%onnx::Conv_790 = Identity(%onnx::Conv_763)
%onnx::Conv_787 = Identity(%onnx::Conv_775)
%onnx::Conv_784 = Identity(%onnx::Conv_775)
%onnx::Conv_781 = Identity(%onnx::Conv_763)
%onnx::Conv_778 = Identity(%onnx::Conv_775)
%onnx::Conv_772 = Identity(%onnx::Conv_763)
%onnx::Conv_769 = Identity(%onnx::Conv_763)
%onnx::Conv_766 = Identity(%onnx::Conv_763)
%onnx::Conv_760 = Identity(%onnx::Conv_733)
%onnx::Conv_757 = Identity(%onnx::Conv_733)
%onnx::Conv_754 = Identity(%onnx::Conv_733)
%onnx::Conv_751 = Identity(%onnx::Conv_733)
%onnx::Conv_748 = Identity(%onnx::Conv_733)
%onnx::Conv_745 = Identity(%onnx::Conv_733)
%onnx::Conv_742 = Identity(%onnx::Conv_733)
%onnx::Conv_739 = Identity(%onnx::Conv_736)
%onnx::Conv_730 = Identity(%onnx::Conv_673)
%onnx::Conv_727 = Identity(%onnx::Conv_673)
%onnx::Conv_724 = Identity(%onnx::Conv_706)
%onnx::Conv_721 = Identity(%onnx::Conv_706)
%onnx::Conv_718 = Identity(%onnx::Conv_706)
%onnx::Conv_715 = Identity(%onnx::Conv_706)
%onnx::Conv_712 = Identity(%onnx::Conv_673)
%onnx::Conv_709 = Identity(%onnx::Conv_673)
%onnx::Conv_703 = Identity(%onnx::Conv_679)
%onnx::Conv_700 = Identity(%onnx::Conv_679)
%onnx::Conv_697 = Identity(%onnx::Conv_679)
%onnx::Conv_694 = Identity(%onnx::Conv_679)
%onnx::Conv_691 = Identity(%onnx::Conv_679)
%onnx::Conv_688 = Identity(%onnx::Conv_679)
%onnx::Conv_685 = Identity(%onnx::Conv_679)
%onnx::Conv_682 = Identity(%onnx::Conv_679)
%onnx::Conv_676 = Identity(%onnx::Conv_673)
%onnx::Conv_670 = Identity(%onnx::Conv_661)
%onnx::Conv_667 = Identity(%onnx::Conv_661)
%onnx::Conv_664 = Identity(%onnx::Conv_661)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_660, %onnx::Conv_661)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_output_0 = Reshape(%/cells.5/nl/Relu_output_0, %/cells.5/shuffle/Constant_output_0)
%/cells.5/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.5/shuffle/Reshape_output_0)
%/cells.5/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.5/shuffle/Reshape_1_output_0 = Reshape(%/cells.5/shuffle/Transpose_output_0, %/cells.5/shuffle/Constant_1_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/shuffle/Reshape_1_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.13/relu/Relu_output_0 = Relu(%/cells.13/conv/Conv_output_0)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/relu/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/relu/Relu_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%658 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %658
}
|
val_accuracy
| 0
| 63,967,360
| 2,337,188
|
{'zcp_synflow': 72.57462921370758, 'zcp_zen': 64.93486785888672, 'zcp_epe_nas': 23.59678698195283, 'zcp_fisher': 0.10879514366388321, 'zcp_flops': 63967360.0, 'zcp_grad_norm': 24.697410583496094, 'zcp_grasp': -0.059307098388671875, 'zcp_jacov': -16.04997194829822, 'zcp_l2_norm': 624.8301391601562, 'zcp_nwot': 207.56905305159125, 'zcp_params': 2337188.0, 'zcp_plain': 0.003105815965682268, 'zcp_snip': 36.83031463623047, 'lat_1080ti_1': 0.5187934591388182, 'lat_1080ti_32': 0.43929510857843135, 'lat_1080ti_64': 0.28259925381327905, 'lat_2080ti_1': 0.6011105610596136, 'lat_2080ti_32': 0.5098005812384533, 'lat_2080ti_64': 0.31302428948303934, 'lat_essential_ph_1': 0.32075471698113206, 'lat_eyeriss': 0.4002989015825281, 'lat_fpga': 0.4464322821322197, 'lat_gold_6226': 0.4381483613295172, 'lat_gold_6240': 0.6099991747351619, 'lat_pixel2': 0.32608695652173914, 'lat_pixel3': 0.380059701402719, 'lat_raspi4': 0.44715533471547514, 'lat_samsung_a50': 0.21052631578947367, 'lat_samsung_s7': 0.2047244094488189, 'lat_silver_4114': 0.6297088404598695, 'lat_silver_4210r': 0.6593218563865639, 'lat_titan_rtx_1': 0.5475993143749995, 'lat_titan_rtx_32': 0.499713394495414, 'lat_titan_rtx_64': 0.35347888606770195, 'lat_titanx_1': 0.2970188732675612, 'lat_titanx_32': 0.45436811919624226, 'lat_titanx_64': 0.33788618482948296, 'lat_titanxp_1': 0.5310449508039465, 'lat_titanxp_32': 0.4558905109934822, 'lat_titanxp_64': 0.34106523291165575}
| |
FBNet_577
|
FBNet
|
577
|
577
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_715[FLOAT, 16x3x3x3]
%onnx::Conv_716[FLOAT, 16]
%onnx::Conv_718[FLOAT, 16x8x1x1]
%onnx::Conv_721[FLOAT, 16x1x5x5]
%onnx::Conv_724[FLOAT, 16x8x1x1]
%onnx::Conv_727[FLOAT, 48x16x1x1]
%onnx::Conv_728[FLOAT, 48]
%onnx::Conv_730[FLOAT, 48x1x3x3]
%onnx::Conv_733[FLOAT, 24x48x1x1]
%onnx::Conv_734[FLOAT, 24]
%onnx::Conv_736[FLOAT, 24x12x1x1]
%onnx::Conv_739[FLOAT, 24x1x3x3]
%onnx::Conv_742[FLOAT, 24x12x1x1]
%onnx::Conv_745[FLOAT, 72x24x1x1]
%onnx::Conv_746[FLOAT, 72]
%onnx::Conv_748[FLOAT, 72x1x3x3]
%onnx::Conv_751[FLOAT, 24x72x1x1]
%onnx::Conv_754[FLOAT, 144x24x1x1]
%onnx::Conv_755[FLOAT, 144]
%onnx::Conv_757[FLOAT, 144x1x3x3]
%onnx::Conv_760[FLOAT, 24x144x1x1]
%onnx::Conv_763[FLOAT, 32x24x1x1]
%onnx::Conv_764[FLOAT, 32]
%onnx::Conv_766[FLOAT, 32x32x1x1]
%onnx::Conv_769[FLOAT, 32x1x5x5]
%onnx::Conv_772[FLOAT, 32x32x1x1]
%onnx::Conv_775[FLOAT, 32x16x1x1]
%onnx::Conv_778[FLOAT, 32x1x3x3]
%onnx::Conv_781[FLOAT, 32x16x1x1]
%onnx::Conv_784[FLOAT, 96x32x1x1]
%onnx::Conv_785[FLOAT, 96]
%onnx::Conv_787[FLOAT, 96x1x5x5]
%onnx::Conv_790[FLOAT, 32x96x1x1]
%onnx::Conv_793[FLOAT, 32x32x1x1]
%onnx::Conv_796[FLOAT, 32x1x5x5]
%onnx::Conv_799[FLOAT, 64x32x1x1]
%onnx::Conv_800[FLOAT, 64]
%onnx::Conv_802[FLOAT, 192x64x1x1]
%onnx::Conv_803[FLOAT, 192]
%onnx::Conv_805[FLOAT, 192x1x3x3]
%onnx::Conv_808[FLOAT, 64x192x1x1]
%onnx::Conv_811[FLOAT, 384x64x1x1]
%onnx::Conv_812[FLOAT, 384]
%onnx::Conv_814[FLOAT, 384x1x5x5]
%onnx::Conv_817[FLOAT, 64x384x1x1]
%onnx::Conv_820[FLOAT, 384x64x1x1]
%onnx::Conv_823[FLOAT, 384x1x3x3]
%onnx::Conv_826[FLOAT, 64x384x1x1]
%onnx::Conv_829[FLOAT, 64x32x1x1]
%onnx::Conv_832[FLOAT, 64x1x3x3]
%onnx::Conv_835[FLOAT, 112x32x1x1]
%onnx::Conv_836[FLOAT, 112]
%onnx::Conv_838[FLOAT, 336x112x1x1]
%onnx::Conv_839[FLOAT, 336]
%onnx::Conv_841[FLOAT, 336x1x3x3]
%onnx::Conv_844[FLOAT, 112x336x1x1]
%onnx::Conv_847[FLOAT, 112x112x1x1]
%onnx::Conv_850[FLOAT, 112x1x5x5]
%onnx::Conv_853[FLOAT, 112x112x1x1]
%onnx::Conv_856[FLOAT, 336x112x1x1]
%onnx::Conv_859[FLOAT, 336x1x3x3]
%onnx::Conv_862[FLOAT, 112x336x1x1]
%onnx::Conv_865[FLOAT, 672x112x1x1]
%onnx::Conv_866[FLOAT, 672]
%onnx::Conv_868[FLOAT, 672x1x5x5]
%onnx::Conv_871[FLOAT, 184x672x1x1]
%onnx::Conv_872[FLOAT, 184]
%onnx::Conv_874[FLOAT, 184x184x1x1]
%onnx::Conv_877[FLOAT, 184x1x3x3]
%onnx::Conv_880[FLOAT, 184x184x1x1]
%onnx::Conv_883[FLOAT, 184x184x1x1]
%onnx::Conv_886[FLOAT, 184x1x3x3]
%onnx::Conv_889[FLOAT, 184x184x1x1]
%onnx::Conv_892[FLOAT, 184x92x1x1]
%onnx::Conv_895[FLOAT, 184x1x5x5]
%onnx::Conv_898[FLOAT, 184x92x1x1]
%onnx::Conv_901[FLOAT, 184x92x1x1]
%onnx::Conv_904[FLOAT, 184x1x3x3]
%onnx::Conv_907[FLOAT, 352x92x1x1]
%onnx::Conv_908[FLOAT, 352]
%onnx::Conv_910[FLOAT, 1504x352x1x1]
%onnx::Conv_911[FLOAT, 1504]
) {
%onnx::Conv_905 = Identity(%onnx::Conv_872)
%onnx::Conv_902 = Identity(%onnx::Conv_872)
%onnx::Conv_899 = Identity(%onnx::Conv_872)
%onnx::Conv_896 = Identity(%onnx::Conv_872)
%onnx::Conv_893 = Identity(%onnx::Conv_872)
%onnx::Conv_890 = Identity(%onnx::Conv_872)
%onnx::Conv_887 = Identity(%onnx::Conv_872)
%onnx::Conv_884 = Identity(%onnx::Conv_872)
%onnx::Conv_881 = Identity(%onnx::Conv_872)
%onnx::Conv_878 = Identity(%onnx::Conv_872)
%onnx::Conv_875 = Identity(%onnx::Conv_872)
%onnx::Conv_869 = Identity(%onnx::Conv_866)
%onnx::Conv_863 = Identity(%onnx::Conv_836)
%onnx::Conv_860 = Identity(%onnx::Conv_839)
%onnx::Conv_857 = Identity(%onnx::Conv_839)
%onnx::Conv_854 = Identity(%onnx::Conv_836)
%onnx::Conv_851 = Identity(%onnx::Conv_836)
%onnx::Conv_848 = Identity(%onnx::Conv_836)
%onnx::Conv_845 = Identity(%onnx::Conv_836)
%onnx::Conv_842 = Identity(%onnx::Conv_839)
%onnx::Conv_833 = Identity(%onnx::Conv_800)
%onnx::Conv_830 = Identity(%onnx::Conv_800)
%onnx::Conv_827 = Identity(%onnx::Conv_800)
%onnx::Conv_824 = Identity(%onnx::Conv_812)
%onnx::Conv_821 = Identity(%onnx::Conv_812)
%onnx::Conv_818 = Identity(%onnx::Conv_800)
%onnx::Conv_815 = Identity(%onnx::Conv_812)
%onnx::Conv_809 = Identity(%onnx::Conv_800)
%onnx::Conv_806 = Identity(%onnx::Conv_803)
%onnx::Conv_797 = Identity(%onnx::Conv_764)
%onnx::Conv_794 = Identity(%onnx::Conv_764)
%onnx::Conv_791 = Identity(%onnx::Conv_764)
%onnx::Conv_788 = Identity(%onnx::Conv_785)
%onnx::Conv_782 = Identity(%onnx::Conv_764)
%onnx::Conv_779 = Identity(%onnx::Conv_764)
%onnx::Conv_776 = Identity(%onnx::Conv_764)
%onnx::Conv_773 = Identity(%onnx::Conv_764)
%onnx::Conv_770 = Identity(%onnx::Conv_764)
%onnx::Conv_767 = Identity(%onnx::Conv_764)
%onnx::Conv_761 = Identity(%onnx::Conv_734)
%onnx::Conv_758 = Identity(%onnx::Conv_755)
%onnx::Conv_752 = Identity(%onnx::Conv_734)
%onnx::Conv_749 = Identity(%onnx::Conv_746)
%onnx::Conv_743 = Identity(%onnx::Conv_734)
%onnx::Conv_740 = Identity(%onnx::Conv_734)
%onnx::Conv_737 = Identity(%onnx::Conv_734)
%onnx::Conv_731 = Identity(%onnx::Conv_728)
%onnx::Conv_725 = Identity(%onnx::Conv_716)
%onnx::Conv_722 = Identity(%onnx::Conv_716)
%onnx::Conv_719 = Identity(%onnx::Conv_716)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_715, %onnx::Conv_716)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_718, %onnx::Conv_719)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_output_0 = Reshape(%/cells.0/nl/Relu_output_0, %/cells.0/shuffle/Constant_output_0)
%/cells.0/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.0/shuffle/Reshape_output_0)
%/cells.0/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.0/shuffle/Reshape_1_output_0 = Reshape(%/cells.0/shuffle/Transpose_output_0, %/cells.0/shuffle/Constant_1_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/shuffle/Reshape_1_output_0, %onnx::Conv_721, %onnx::Conv_722)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_724, %onnx::Conv_725)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_727, %onnx::Conv_728)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_730, %onnx::Conv_731)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_733, %onnx::Conv_734)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_736, %onnx::Conv_737)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_739, %onnx::Conv_740)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_742, %onnx::Conv_743)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_745, %onnx::Conv_746)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_748, %onnx::Conv_749)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_751, %onnx::Conv_752)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_754, %onnx::Conv_755)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_757, %onnx::Conv_758)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_760, %onnx::Conv_761)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.4/Add_output_0, %onnx::Conv_763, %onnx::Conv_764)
%/cells.5/relu/Relu_output_0 = Relu(%/cells.5/conv/Conv_output_0)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/relu/Relu_output_0, %onnx::Conv_766, %onnx::Conv_767)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/nl/Relu_output_0, %onnx::Conv_769, %onnx::Conv_770)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_772, %onnx::Conv_773)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/relu/Relu_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_775, %onnx::Conv_776)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_output_0 = Reshape(%/cells.7/nl/Relu_output_0, %/cells.7/shuffle/Constant_output_0)
%/cells.7/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.7/shuffle/Reshape_output_0)
%/cells.7/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.7/shuffle/Reshape_1_output_0 = Reshape(%/cells.7/shuffle/Transpose_output_0, %/cells.7/shuffle/Constant_1_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/shuffle/Reshape_1_output_0, %onnx::Conv_778, %onnx::Conv_779)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_781, %onnx::Conv_782)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_784, %onnx::Conv_785)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_787, %onnx::Conv_788)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_790, %onnx::Conv_791)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_793, %onnx::Conv_794)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_796, %onnx::Conv_797)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_799, %onnx::Conv_800)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_802, %onnx::Conv_803)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_805, %onnx::Conv_806)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_808, %onnx::Conv_809)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_811, %onnx::Conv_812)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_814, %onnx::Conv_815)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_817, %onnx::Conv_818)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_820, %onnx::Conv_821)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_823, %onnx::Conv_824)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_826, %onnx::Conv_827)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_829, %onnx::Conv_830)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_output_0 = Reshape(%/cells.13/nl/Relu_output_0, %/cells.13/shuffle/Constant_output_0)
%/cells.13/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.13/shuffle/Reshape_output_0)
%/cells.13/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.13/shuffle/Reshape_1_output_0 = Reshape(%/cells.13/shuffle/Transpose_output_0, %/cells.13/shuffle/Constant_1_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/shuffle/Reshape_1_output_0, %onnx::Conv_832, %onnx::Conv_833)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_835, %onnx::Conv_836)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_838, %onnx::Conv_839)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_841, %onnx::Conv_842)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_844, %onnx::Conv_845)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_847, %onnx::Conv_848)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_850, %onnx::Conv_851)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_853, %onnx::Conv_854)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_856, %onnx::Conv_857)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_859, %onnx::Conv_860)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_862, %onnx::Conv_863)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_865, %onnx::Conv_866)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_868, %onnx::Conv_869)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_871, %onnx::Conv_872)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_874, %onnx::Conv_875)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_877, %onnx::Conv_878)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_880, %onnx::Conv_881)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_883, %onnx::Conv_884)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_886, %onnx::Conv_887)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_889, %onnx::Conv_890)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_892, %onnx::Conv_893)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_output_0 = Reshape(%/cells.20/nl/Relu_output_0, %/cells.20/shuffle/Constant_output_0)
%/cells.20/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.20/shuffle/Reshape_output_0)
%/cells.20/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.20/shuffle/Reshape_1_output_0 = Reshape(%/cells.20/shuffle/Transpose_output_0, %/cells.20/shuffle/Constant_1_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.20/shuffle/Reshape_1_output_0, %onnx::Conv_895, %onnx::Conv_896)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_898, %onnx::Conv_899)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_901, %onnx::Conv_902)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_904, %onnx::Conv_905)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_907, %onnx::Conv_908)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_910, %onnx::Conv_911)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%713 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %713
}
|
val_accuracy
| 0
| 64,296,320
| 1,507,956
|
{'zcp_synflow': 79.27452925517403, 'zcp_zen': 69.97154235839844, 'zcp_epe_nas': 10.342585441763173, 'zcp_fisher': 0.18924430012702942, 'zcp_flops': 64296320.0, 'zcp_grad_norm': 26.617036819458008, 'zcp_grasp': 0.0064449310302734375, 'zcp_jacov': -16.05157838879841, 'zcp_l2_norm': 627.1903076171875, 'zcp_nwot': 212.46149206602604, 'zcp_params': 1507956.0, 'zcp_plain': -0.002724145771935582, 'zcp_snip': 40.30386734008789, 'lat_1080ti_1': 0.704859998660373, 'lat_1080ti_32': 0.6739135334443838, 'lat_1080ti_64': 0.49540970253788014, 'lat_2080ti_1': 0.7568068817115161, 'lat_2080ti_32': 0.6933089439203938, 'lat_2080ti_64': 0.5077989423918451, 'lat_essential_ph_1': 0.37735849056603776, 'lat_eyeriss': 0.3771844571177087, 'lat_fpga': 0.41083921273711393, 'lat_gold_6226': 0.2609121193116213, 'lat_gold_6240': 0.46568130217116166, 'lat_pixel2': 0.2608695652173913, 'lat_pixel3': 0.3564388932131067, 'lat_raspi4': 0.3487101984199767, 'lat_samsung_a50': 0.16842105263157894, 'lat_samsung_s7': 0.2047244094488189, 'lat_silver_4114': 0.49812947593739265, 'lat_silver_4210r': 0.519228294428724, 'lat_titan_rtx_1': 0.7448749060131399, 'lat_titan_rtx_32': 0.6753592145277043, 'lat_titan_rtx_64': 0.561893815356066, 'lat_titanx_1': 0.3900642391231552, 'lat_titanx_32': 0.5863748275306547, 'lat_titanx_64': 0.45561797199178583, 'lat_titanxp_1': 0.6735088760627007, 'lat_titanxp_32': 0.6591553255063316, 'lat_titanxp_64': 0.5099376300396621}
| |
FBNet_3469
|
FBNet
|
3469
|
3469
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_732[FLOAT, 16x3x3x3]
%onnx::Conv_733[FLOAT, 16]
%onnx::Conv_735[FLOAT, 48x16x1x1]
%onnx::Conv_736[FLOAT, 48]
%onnx::Conv_738[FLOAT, 48x1x5x5]
%onnx::Conv_741[FLOAT, 16x48x1x1]
%onnx::Conv_744[FLOAT, 16x8x1x1]
%onnx::Conv_747[FLOAT, 16x1x3x3]
%onnx::Conv_750[FLOAT, 24x8x1x1]
%onnx::Conv_751[FLOAT, 24]
%onnx::Conv_753[FLOAT, 24x24x1x1]
%onnx::Conv_756[FLOAT, 24x1x3x3]
%onnx::Conv_759[FLOAT, 24x24x1x1]
%onnx::Conv_762[FLOAT, 144x24x1x1]
%onnx::Conv_763[FLOAT, 144]
%onnx::Conv_765[FLOAT, 144x1x5x5]
%onnx::Conv_768[FLOAT, 24x144x1x1]
%onnx::Conv_771[FLOAT, 144x24x1x1]
%onnx::Conv_774[FLOAT, 144x1x3x3]
%onnx::Conv_777[FLOAT, 24x144x1x1]
%onnx::Conv_780[FLOAT, 24x24x1x1]
%onnx::Conv_783[FLOAT, 24x1x3x3]
%onnx::Conv_786[FLOAT, 32x24x1x1]
%onnx::Conv_787[FLOAT, 32]
%onnx::Conv_789[FLOAT, 32x16x1x1]
%onnx::Conv_792[FLOAT, 32x1x3x3]
%onnx::Conv_795[FLOAT, 32x16x1x1]
%onnx::Conv_798[FLOAT, 32x32x1x1]
%onnx::Conv_801[FLOAT, 32x1x3x3]
%onnx::Conv_804[FLOAT, 32x32x1x1]
%onnx::Conv_807[FLOAT, 192x32x1x1]
%onnx::Conv_808[FLOAT, 192]
%onnx::Conv_810[FLOAT, 192x1x5x5]
%onnx::Conv_813[FLOAT, 32x192x1x1]
%onnx::Conv_816[FLOAT, 96x32x1x1]
%onnx::Conv_817[FLOAT, 96]
%onnx::Conv_819[FLOAT, 96x1x5x5]
%onnx::Conv_822[FLOAT, 64x96x1x1]
%onnx::Conv_823[FLOAT, 64]
%onnx::Conv_825[FLOAT, 384x64x1x1]
%onnx::Conv_826[FLOAT, 384]
%onnx::Conv_828[FLOAT, 384x1x5x5]
%onnx::Conv_831[FLOAT, 64x384x1x1]
%onnx::Conv_834[FLOAT, 384x64x1x1]
%onnx::Conv_837[FLOAT, 384x1x3x3]
%onnx::Conv_840[FLOAT, 64x384x1x1]
%onnx::Conv_843[FLOAT, 64x32x1x1]
%onnx::Conv_846[FLOAT, 64x1x5x5]
%onnx::Conv_849[FLOAT, 64x32x1x1]
%onnx::Conv_852[FLOAT, 384x64x1x1]
%onnx::Conv_855[FLOAT, 384x1x3x3]
%onnx::Conv_858[FLOAT, 112x384x1x1]
%onnx::Conv_859[FLOAT, 112]
%onnx::Conv_861[FLOAT, 336x112x1x1]
%onnx::Conv_862[FLOAT, 336]
%onnx::Conv_864[FLOAT, 336x1x5x5]
%onnx::Conv_867[FLOAT, 112x336x1x1]
%onnx::Conv_870[FLOAT, 112x112x1x1]
%onnx::Conv_873[FLOAT, 112x1x3x3]
%onnx::Conv_876[FLOAT, 112x112x1x1]
%onnx::Conv_879[FLOAT, 336x112x1x1]
%onnx::Conv_882[FLOAT, 336x1x3x3]
%onnx::Conv_885[FLOAT, 112x336x1x1]
%onnx::Conv_888[FLOAT, 112x56x1x1]
%onnx::Conv_891[FLOAT, 112x1x5x5]
%onnx::Conv_894[FLOAT, 184x56x1x1]
%onnx::Conv_895[FLOAT, 184]
%onnx::Conv_897[FLOAT, 184x92x1x1]
%onnx::Conv_900[FLOAT, 184x1x3x3]
%onnx::Conv_903[FLOAT, 184x92x1x1]
%onnx::Conv_906[FLOAT, 184x92x1x1]
%onnx::Conv_909[FLOAT, 184x1x3x3]
%onnx::Conv_912[FLOAT, 184x92x1x1]
%onnx::Conv_915[FLOAT, 184x184x1x1]
%onnx::Conv_918[FLOAT, 184x1x3x3]
%onnx::Conv_921[FLOAT, 184x184x1x1]
%onnx::Conv_924[FLOAT, 184x184x1x1]
%onnx::Conv_927[FLOAT, 184x1x3x3]
%onnx::Conv_930[FLOAT, 352x184x1x1]
%onnx::Conv_931[FLOAT, 352]
%onnx::Conv_933[FLOAT, 1504x352x1x1]
%onnx::Conv_934[FLOAT, 1504]
) {
%onnx::Conv_928 = Identity(%onnx::Conv_895)
%onnx::Conv_925 = Identity(%onnx::Conv_895)
%onnx::Conv_922 = Identity(%onnx::Conv_895)
%onnx::Conv_919 = Identity(%onnx::Conv_895)
%onnx::Conv_916 = Identity(%onnx::Conv_895)
%onnx::Conv_913 = Identity(%onnx::Conv_895)
%onnx::Conv_910 = Identity(%onnx::Conv_895)
%onnx::Conv_907 = Identity(%onnx::Conv_895)
%onnx::Conv_904 = Identity(%onnx::Conv_895)
%onnx::Conv_901 = Identity(%onnx::Conv_895)
%onnx::Conv_898 = Identity(%onnx::Conv_895)
%onnx::Conv_892 = Identity(%onnx::Conv_859)
%onnx::Conv_889 = Identity(%onnx::Conv_859)
%onnx::Conv_886 = Identity(%onnx::Conv_859)
%onnx::Conv_883 = Identity(%onnx::Conv_862)
%onnx::Conv_880 = Identity(%onnx::Conv_862)
%onnx::Conv_877 = Identity(%onnx::Conv_859)
%onnx::Conv_874 = Identity(%onnx::Conv_859)
%onnx::Conv_871 = Identity(%onnx::Conv_859)
%onnx::Conv_868 = Identity(%onnx::Conv_859)
%onnx::Conv_865 = Identity(%onnx::Conv_862)
%onnx::Conv_856 = Identity(%onnx::Conv_826)
%onnx::Conv_853 = Identity(%onnx::Conv_826)
%onnx::Conv_850 = Identity(%onnx::Conv_823)
%onnx::Conv_847 = Identity(%onnx::Conv_823)
%onnx::Conv_844 = Identity(%onnx::Conv_823)
%onnx::Conv_841 = Identity(%onnx::Conv_823)
%onnx::Conv_838 = Identity(%onnx::Conv_826)
%onnx::Conv_835 = Identity(%onnx::Conv_826)
%onnx::Conv_832 = Identity(%onnx::Conv_823)
%onnx::Conv_829 = Identity(%onnx::Conv_826)
%onnx::Conv_820 = Identity(%onnx::Conv_817)
%onnx::Conv_814 = Identity(%onnx::Conv_787)
%onnx::Conv_811 = Identity(%onnx::Conv_808)
%onnx::Conv_805 = Identity(%onnx::Conv_787)
%onnx::Conv_802 = Identity(%onnx::Conv_787)
%onnx::Conv_799 = Identity(%onnx::Conv_787)
%onnx::Conv_796 = Identity(%onnx::Conv_787)
%onnx::Conv_793 = Identity(%onnx::Conv_787)
%onnx::Conv_790 = Identity(%onnx::Conv_787)
%onnx::Conv_784 = Identity(%onnx::Conv_751)
%onnx::Conv_781 = Identity(%onnx::Conv_751)
%onnx::Conv_778 = Identity(%onnx::Conv_751)
%onnx::Conv_775 = Identity(%onnx::Conv_763)
%onnx::Conv_772 = Identity(%onnx::Conv_763)
%onnx::Conv_769 = Identity(%onnx::Conv_751)
%onnx::Conv_766 = Identity(%onnx::Conv_763)
%onnx::Conv_760 = Identity(%onnx::Conv_751)
%onnx::Conv_757 = Identity(%onnx::Conv_751)
%onnx::Conv_754 = Identity(%onnx::Conv_751)
%onnx::Conv_748 = Identity(%onnx::Conv_733)
%onnx::Conv_745 = Identity(%onnx::Conv_733)
%onnx::Conv_742 = Identity(%onnx::Conv_733)
%onnx::Conv_739 = Identity(%onnx::Conv_736)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_732, %onnx::Conv_733)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_output_0 = Reshape(%/cells.1/nl/Relu_output_0, %/cells.1/shuffle/Constant_output_0)
%/cells.1/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.1/shuffle/Reshape_output_0)
%/cells.1/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.1/shuffle/Reshape_1_output_0 = Reshape(%/cells.1/shuffle/Transpose_output_0, %/cells.1/shuffle/Constant_1_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 16, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/shuffle/Reshape_1_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_840, %onnx::Conv_841)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_843, %onnx::Conv_844)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_output_0 = Reshape(%/cells.12/nl/Relu_output_0, %/cells.12/shuffle/Constant_output_0)
%/cells.12/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.12/shuffle/Reshape_output_0)
%/cells.12/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.12/shuffle/Reshape_1_output_0 = Reshape(%/cells.12/shuffle/Transpose_output_0, %/cells.12/shuffle/Constant_1_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/shuffle/Reshape_1_output_0, %onnx::Conv_846, %onnx::Conv_847)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_849, %onnx::Conv_850)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_852, %onnx::Conv_853)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 384, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_855, %onnx::Conv_856)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_858, %onnx::Conv_859)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_861, %onnx::Conv_862)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_864, %onnx::Conv_865)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_867, %onnx::Conv_868)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_870, %onnx::Conv_871)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_873, %onnx::Conv_874)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_876, %onnx::Conv_877)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_879, %onnx::Conv_880)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_882, %onnx::Conv_883)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_885, %onnx::Conv_886)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_888, %onnx::Conv_889)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_output_0 = Reshape(%/cells.17/nl/Relu_output_0, %/cells.17/shuffle/Constant_output_0)
%/cells.17/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.17/shuffle/Reshape_output_0)
%/cells.17/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.17/shuffle/Reshape_1_output_0 = Reshape(%/cells.17/shuffle/Transpose_output_0, %/cells.17/shuffle/Constant_1_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/shuffle/Reshape_1_output_0, %onnx::Conv_891, %onnx::Conv_892)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_894, %onnx::Conv_895)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_897, %onnx::Conv_898)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_output_0 = Reshape(%/cells.18/nl/Relu_output_0, %/cells.18/shuffle/Constant_output_0)
%/cells.18/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.18/shuffle/Reshape_output_0)
%/cells.18/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.18/shuffle/Reshape_1_output_0 = Reshape(%/cells.18/shuffle/Transpose_output_0, %/cells.18/shuffle/Constant_1_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/shuffle/Reshape_1_output_0, %onnx::Conv_900, %onnx::Conv_901)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_903, %onnx::Conv_904)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_906, %onnx::Conv_907)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_output_0 = Reshape(%/cells.19/nl/Relu_output_0, %/cells.19/shuffle/Constant_output_0)
%/cells.19/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.19/shuffle/Reshape_output_0)
%/cells.19/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.19/shuffle/Reshape_1_output_0 = Reshape(%/cells.19/shuffle/Transpose_output_0, %/cells.19/shuffle/Constant_1_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/shuffle/Reshape_1_output_0, %onnx::Conv_909, %onnx::Conv_910)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_912, %onnx::Conv_913)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_915, %onnx::Conv_916)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_918, %onnx::Conv_919)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_921, %onnx::Conv_922)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_924, %onnx::Conv_925)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.21/nl/Relu_output_0, %onnx::Conv_927, %onnx::Conv_928)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_930, %onnx::Conv_931)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_933, %onnx::Conv_934)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%730 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %730
}
|
val_accuracy
| 0
| 72,600,448
| 1,395,236
|
{'zcp_synflow': 80.83160453771569, 'zcp_zen': 71.71893310546875, 'zcp_epe_nas': 9.471443902087355, 'zcp_fisher': 0.18494202196598053, 'zcp_flops': 72600448.0, 'zcp_grad_norm': 33.09793472290039, 'zcp_grasp': -0.020465850830078125, 'zcp_jacov': -16.058990506305737, 'zcp_l2_norm': 632.27734375, 'zcp_nwot': 215.58835135832712, 'zcp_params': 1395236.0, 'zcp_plain': -0.0019171343883499503, 'zcp_snip': 54.15190505981445, 'lat_1080ti_1': 0.8322049219880583, 'lat_1080ti_32': 0.8414451917124169, 'lat_1080ti_64': 0.6753226350274353, 'lat_2080ti_1': 0.8444871985124607, 'lat_2080ti_32': 0.860852508148026, 'lat_2080ti_64': 0.7211760241226172, 'lat_essential_ph_1': 0.32075471698113206, 'lat_eyeriss': 0.5098014046540627, 'lat_fpga': 0.5127400825260491, 'lat_gold_6226': 0.28025302984334927, 'lat_gold_6240': 0.5016060455307761, 'lat_pixel2': 0.30434782608695654, 'lat_pixel3': 0.5275816820050628, 'lat_raspi4': 0.4967252947736609, 'lat_samsung_a50': 0.2631578947368421, 'lat_samsung_s7': 0.1732283464566929, 'lat_silver_4114': 0.544448207707855, 'lat_silver_4210r': 0.5951442042536238, 'lat_titan_rtx_1': 0.801922597994827, 'lat_titan_rtx_32': 0.8313275753470536, 'lat_titan_rtx_64': 0.778414416172408, 'lat_titanx_1': 0.4310567274189119, 'lat_titanx_32': 0.8146467671289058, 'lat_titanx_64': 0.6649151476414454, 'lat_titanxp_1': 0.7710718417542466, 'lat_titanxp_32': 0.8236429880682157, 'lat_titanxp_64': 0.7237191764707877}
| |
FBNet_4589
|
FBNet
|
4589
|
4589
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_615[FLOAT, 16x3x3x3]
%onnx::Conv_616[FLOAT, 16]
%onnx::Conv_618[FLOAT, 48x16x1x1]
%onnx::Conv_619[FLOAT, 48]
%onnx::Conv_621[FLOAT, 48x1x3x3]
%onnx::Conv_624[FLOAT, 16x48x1x1]
%onnx::Conv_627[FLOAT, 24x16x1x1]
%onnx::Conv_628[FLOAT, 24]
%onnx::Conv_630[FLOAT, 24x24x1x1]
%onnx::Conv_633[FLOAT, 24x1x3x3]
%onnx::Conv_636[FLOAT, 24x24x1x1]
%onnx::Conv_639[FLOAT, 24x12x1x1]
%onnx::Conv_642[FLOAT, 24x1x3x3]
%onnx::Conv_645[FLOAT, 24x12x1x1]
%onnx::Conv_648[FLOAT, 72x24x1x1]
%onnx::Conv_649[FLOAT, 72]
%onnx::Conv_651[FLOAT, 72x1x3x3]
%onnx::Conv_654[FLOAT, 32x72x1x1]
%onnx::Conv_655[FLOAT, 32]
%onnx::Conv_657[FLOAT, 32x16x1x1]
%onnx::Conv_660[FLOAT, 32x1x5x5]
%onnx::Conv_663[FLOAT, 32x16x1x1]
%onnx::Conv_666[FLOAT, 192x32x1x1]
%onnx::Conv_667[FLOAT, 192]
%onnx::Conv_669[FLOAT, 192x1x3x3]
%onnx::Conv_672[FLOAT, 32x192x1x1]
%onnx::Conv_675[FLOAT, 32x32x1x1]
%onnx::Conv_678[FLOAT, 32x1x5x5]
%onnx::Conv_681[FLOAT, 32x32x1x1]
%onnx::Conv_684[FLOAT, 96x32x1x1]
%onnx::Conv_685[FLOAT, 96]
%onnx::Conv_687[FLOAT, 96x1x3x3]
%onnx::Conv_690[FLOAT, 64x96x1x1]
%onnx::Conv_691[FLOAT, 64]
%onnx::Conv_693[FLOAT, 192x64x1x1]
%onnx::Conv_696[FLOAT, 192x1x5x5]
%onnx::Conv_699[FLOAT, 64x192x1x1]
%onnx::Conv_702[FLOAT, 64x64x1x1]
%onnx::Conv_705[FLOAT, 64x1x5x5]
%onnx::Conv_708[FLOAT, 64x64x1x1]
%onnx::Conv_711[FLOAT, 192x64x1x1]
%onnx::Conv_714[FLOAT, 192x1x3x3]
%onnx::Conv_717[FLOAT, 64x192x1x1]
%onnx::Conv_720[FLOAT, 64x64x1x1]
%onnx::Conv_723[FLOAT, 64x1x3x3]
%onnx::Conv_726[FLOAT, 112x64x1x1]
%onnx::Conv_727[FLOAT, 112]
%onnx::Conv_729[FLOAT, 112x56x1x1]
%onnx::Conv_732[FLOAT, 112x1x5x5]
%onnx::Conv_735[FLOAT, 112x56x1x1]
%onnx::Conv_738[FLOAT, 112x112x1x1]
%onnx::Conv_741[FLOAT, 112x1x5x5]
%onnx::Conv_744[FLOAT, 112x112x1x1]
%onnx::Conv_747[FLOAT, 336x112x1x1]
%onnx::Conv_748[FLOAT, 336]
%onnx::Conv_750[FLOAT, 336x1x3x3]
%onnx::Conv_753[FLOAT, 112x336x1x1]
%onnx::Conv_756[FLOAT, 112x112x1x1]
%onnx::Conv_759[FLOAT, 112x1x3x3]
%onnx::Conv_762[FLOAT, 184x112x1x1]
%onnx::Conv_763[FLOAT, 184]
%onnx::Conv_765[FLOAT, 184x184x1x1]
%onnx::Conv_768[FLOAT, 184x1x3x3]
%onnx::Conv_771[FLOAT, 184x184x1x1]
%onnx::Conv_774[FLOAT, 552x184x1x1]
%onnx::Conv_775[FLOAT, 552]
%onnx::Conv_777[FLOAT, 552x1x3x3]
%onnx::Conv_780[FLOAT, 184x552x1x1]
%onnx::Conv_783[FLOAT, 552x184x1x1]
%onnx::Conv_786[FLOAT, 552x1x3x3]
%onnx::Conv_789[FLOAT, 184x552x1x1]
%onnx::Conv_792[FLOAT, 352x184x1x1]
%onnx::Conv_793[FLOAT, 352]
%onnx::Conv_795[FLOAT, 1504x352x1x1]
%onnx::Conv_796[FLOAT, 1504]
) {
%onnx::Conv_790 = Identity(%onnx::Conv_763)
%onnx::Conv_787 = Identity(%onnx::Conv_775)
%onnx::Conv_784 = Identity(%onnx::Conv_775)
%onnx::Conv_781 = Identity(%onnx::Conv_763)
%onnx::Conv_778 = Identity(%onnx::Conv_775)
%onnx::Conv_772 = Identity(%onnx::Conv_763)
%onnx::Conv_769 = Identity(%onnx::Conv_763)
%onnx::Conv_766 = Identity(%onnx::Conv_763)
%onnx::Conv_760 = Identity(%onnx::Conv_727)
%onnx::Conv_757 = Identity(%onnx::Conv_727)
%onnx::Conv_754 = Identity(%onnx::Conv_727)
%onnx::Conv_751 = Identity(%onnx::Conv_748)
%onnx::Conv_745 = Identity(%onnx::Conv_727)
%onnx::Conv_742 = Identity(%onnx::Conv_727)
%onnx::Conv_739 = Identity(%onnx::Conv_727)
%onnx::Conv_736 = Identity(%onnx::Conv_727)
%onnx::Conv_733 = Identity(%onnx::Conv_727)
%onnx::Conv_730 = Identity(%onnx::Conv_727)
%onnx::Conv_724 = Identity(%onnx::Conv_691)
%onnx::Conv_721 = Identity(%onnx::Conv_691)
%onnx::Conv_718 = Identity(%onnx::Conv_691)
%onnx::Conv_715 = Identity(%onnx::Conv_667)
%onnx::Conv_712 = Identity(%onnx::Conv_667)
%onnx::Conv_709 = Identity(%onnx::Conv_691)
%onnx::Conv_706 = Identity(%onnx::Conv_691)
%onnx::Conv_703 = Identity(%onnx::Conv_691)
%onnx::Conv_700 = Identity(%onnx::Conv_691)
%onnx::Conv_697 = Identity(%onnx::Conv_667)
%onnx::Conv_694 = Identity(%onnx::Conv_667)
%onnx::Conv_688 = Identity(%onnx::Conv_685)
%onnx::Conv_682 = Identity(%onnx::Conv_655)
%onnx::Conv_679 = Identity(%onnx::Conv_655)
%onnx::Conv_676 = Identity(%onnx::Conv_655)
%onnx::Conv_673 = Identity(%onnx::Conv_655)
%onnx::Conv_670 = Identity(%onnx::Conv_667)
%onnx::Conv_664 = Identity(%onnx::Conv_655)
%onnx::Conv_661 = Identity(%onnx::Conv_655)
%onnx::Conv_658 = Identity(%onnx::Conv_655)
%onnx::Conv_652 = Identity(%onnx::Conv_649)
%onnx::Conv_646 = Identity(%onnx::Conv_628)
%onnx::Conv_643 = Identity(%onnx::Conv_628)
%onnx::Conv_640 = Identity(%onnx::Conv_628)
%onnx::Conv_637 = Identity(%onnx::Conv_628)
%onnx::Conv_634 = Identity(%onnx::Conv_628)
%onnx::Conv_631 = Identity(%onnx::Conv_628)
%onnx::Conv_625 = Identity(%onnx::Conv_616)
%onnx::Conv_622 = Identity(%onnx::Conv_619)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_615, %onnx::Conv_616)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.0/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_618, %onnx::Conv_619)
%/cells.0/nl/Relu_output_0 = Relu(%/cells.0/conv1/Conv_output_0)
%/cells.0/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.0/nl/Relu_output_0, %onnx::Conv_621, %onnx::Conv_622)
%/cells.0/nl_1/Relu_output_0 = Relu(%/cells.0/conv2/Conv_output_0)
%/cells.0/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/nl_1/Relu_output_0, %onnx::Conv_624, %onnx::Conv_625)
%/cells.0/Add_output_0 = Add(%/cells.0/conv3/Conv_output_0, %/stem/relu/Relu_output_0)
%/cells.1/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.0/Add_output_0, %onnx::Conv_627, %onnx::Conv_628)
%/cells.1/relu/Relu_output_0 = Relu(%/cells.1/conv/Conv_output_0)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/relu/Relu_output_0, %onnx::Conv_630, %onnx::Conv_631)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/nl/Relu_output_0, %onnx::Conv_633, %onnx::Conv_634)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_636, %onnx::Conv_637)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/relu/Relu_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_639, %onnx::Conv_640)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_output_0 = Reshape(%/cells.4/nl/Relu_output_0, %/cells.4/shuffle/Constant_output_0)
%/cells.4/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.4/shuffle/Reshape_output_0)
%/cells.4/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.4/shuffle/Reshape_1_output_0 = Reshape(%/cells.4/shuffle/Transpose_output_0, %/cells.4/shuffle/Constant_1_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.4/shuffle/Reshape_1_output_0, %onnx::Conv_642, %onnx::Conv_643)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_645, %onnx::Conv_646)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_648, %onnx::Conv_649)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_651, %onnx::Conv_652)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/Add_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.9/nl/Relu_output_0 = Relu(%/cells.9/conv1/Conv_output_0)
%/cells.9/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.9/nl/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.9/nl_1/Relu_output_0 = Relu(%/cells.9/conv2/Conv_output_0)
%/cells.9/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/nl_1/Relu_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/conv3/Conv_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/conv3/Conv_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.11/nl/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_output_0 = Reshape(%/cells.14/nl/Relu_output_0, %/cells.14/shuffle/Constant_output_0)
%/cells.14/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.14/shuffle/Reshape_output_0)
%/cells.14/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.14/shuffle/Reshape_1_output_0 = Reshape(%/cells.14/shuffle/Transpose_output_0, %/cells.14/shuffle/Constant_1_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.14/shuffle/Reshape_1_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 336, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 552, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.21/relu/Relu_output_0 = Relu(%/cells.21/conv/Conv_output_0)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/relu/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%613 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %613
}
|
val_accuracy
| 0
| 47,008,896
| 1,520,660
|
{'zcp_synflow': 78.2644170523288, 'zcp_zen': 65.50042724609375, 'zcp_epe_nas': 22.22151084336175, 'zcp_fisher': 0.07629679143428802, 'zcp_flops': 47008896.0, 'zcp_grad_norm': 19.305469512939453, 'zcp_grasp': 0.012579917907714844, 'zcp_jacov': -16.062446604743577, 'zcp_l2_norm': 588.0046997070312, 'zcp_nwot': 206.12957976275922, 'zcp_params': 1520660.0, 'zcp_plain': 0.0007459570188075304, 'zcp_snip': 32.073829650878906, 'lat_1080ti_1': 0.5295520985488632, 'lat_1080ti_32': 0.4572642678520407, 'lat_1080ti_64': 0.19758301914128534, 'lat_2080ti_1': 0.5656716324008045, 'lat_2080ti_32': 0.41206708725065416, 'lat_2080ti_64': 0.2534069801737127, 'lat_essential_ph_1': 0.16981132075471697, 'lat_eyeriss': 0.1773238222727521, 'lat_fpga': 0.18553518241819134, 'lat_gold_6226': 0.18249403567328962, 'lat_gold_6240': 0.273324823398122, 'lat_pixel2': 0.2391304347826087, 'lat_pixel3': 0.14039483155202104, 'lat_raspi4': 0.15575385130001185, 'lat_samsung_a50': 0.08421052631578947, 'lat_samsung_s7': 0.08661417322834646, 'lat_silver_4114': 0.32636833027365847, 'lat_silver_4210r': 0.31679384307780417, 'lat_titan_rtx_1': 0.5260664823246749, 'lat_titan_rtx_32': 0.428095444233819, 'lat_titan_rtx_64': 0.2689013467153393, 'lat_titanx_1': 0.27697326559466323, 'lat_titanx_32': 0.30839623153237467, 'lat_titanx_64': 0.18509874982237284, 'lat_titanxp_1': 0.48851768787759775, 'lat_titanxp_32': 0.36970102316697806, 'lat_titanxp_64': 0.23249871858983637}
| |
FBNet_470
|
FBNet
|
470
|
470
|
graph main_graph (
%input.1[FLOAT, 1x3x32x32]
%fc.weight[FLOAT, 100x1504]
%fc.bias[FLOAT, 100]
%onnx::Conv_651[FLOAT, 16x3x3x3]
%onnx::Conv_652[FLOAT, 16]
%onnx::Conv_654[FLOAT, 48x16x1x1]
%onnx::Conv_655[FLOAT, 48]
%onnx::Conv_657[FLOAT, 48x1x3x3]
%onnx::Conv_660[FLOAT, 24x48x1x1]
%onnx::Conv_661[FLOAT, 24]
%onnx::Conv_663[FLOAT, 24x12x1x1]
%onnx::Conv_666[FLOAT, 24x1x3x3]
%onnx::Conv_669[FLOAT, 24x12x1x1]
%onnx::Conv_672[FLOAT, 144x24x1x1]
%onnx::Conv_673[FLOAT, 144]
%onnx::Conv_675[FLOAT, 144x1x5x5]
%onnx::Conv_678[FLOAT, 24x144x1x1]
%onnx::Conv_681[FLOAT, 72x24x1x1]
%onnx::Conv_682[FLOAT, 72]
%onnx::Conv_684[FLOAT, 72x1x5x5]
%onnx::Conv_687[FLOAT, 24x72x1x1]
%onnx::Conv_690[FLOAT, 24x24x1x1]
%onnx::Conv_693[FLOAT, 24x1x5x5]
%onnx::Conv_696[FLOAT, 32x24x1x1]
%onnx::Conv_697[FLOAT, 32]
%onnx::Conv_699[FLOAT, 32x16x1x1]
%onnx::Conv_702[FLOAT, 32x1x5x5]
%onnx::Conv_705[FLOAT, 32x16x1x1]
%onnx::Conv_708[FLOAT, 192x32x1x1]
%onnx::Conv_709[FLOAT, 192]
%onnx::Conv_711[FLOAT, 192x1x5x5]
%onnx::Conv_714[FLOAT, 32x192x1x1]
%onnx::Conv_717[FLOAT, 96x32x1x1]
%onnx::Conv_718[FLOAT, 96]
%onnx::Conv_720[FLOAT, 96x1x5x5]
%onnx::Conv_723[FLOAT, 32x96x1x1]
%onnx::Conv_726[FLOAT, 64x32x1x1]
%onnx::Conv_727[FLOAT, 64]
%onnx::Conv_729[FLOAT, 192x64x1x1]
%onnx::Conv_732[FLOAT, 192x1x5x5]
%onnx::Conv_735[FLOAT, 64x192x1x1]
%onnx::Conv_738[FLOAT, 64x32x1x1]
%onnx::Conv_741[FLOAT, 64x1x3x3]
%onnx::Conv_744[FLOAT, 64x32x1x1]
%onnx::Conv_747[FLOAT, 192x64x1x1]
%onnx::Conv_750[FLOAT, 192x1x5x5]
%onnx::Conv_753[FLOAT, 64x192x1x1]
%onnx::Conv_756[FLOAT, 64x64x1x1]
%onnx::Conv_759[FLOAT, 64x1x5x5]
%onnx::Conv_762[FLOAT, 112x64x1x1]
%onnx::Conv_763[FLOAT, 112]
%onnx::Conv_765[FLOAT, 672x112x1x1]
%onnx::Conv_766[FLOAT, 672]
%onnx::Conv_768[FLOAT, 672x1x3x3]
%onnx::Conv_771[FLOAT, 112x672x1x1]
%onnx::Conv_774[FLOAT, 672x112x1x1]
%onnx::Conv_777[FLOAT, 672x1x3x3]
%onnx::Conv_780[FLOAT, 112x672x1x1]
%onnx::Conv_783[FLOAT, 672x112x1x1]
%onnx::Conv_786[FLOAT, 672x1x5x5]
%onnx::Conv_789[FLOAT, 112x672x1x1]
%onnx::Conv_792[FLOAT, 112x112x1x1]
%onnx::Conv_795[FLOAT, 112x1x5x5]
%onnx::Conv_798[FLOAT, 184x112x1x1]
%onnx::Conv_799[FLOAT, 184]
%onnx::Conv_801[FLOAT, 1104x184x1x1]
%onnx::Conv_802[FLOAT, 1104]
%onnx::Conv_804[FLOAT, 1104x1x5x5]
%onnx::Conv_807[FLOAT, 184x1104x1x1]
%onnx::Conv_810[FLOAT, 184x184x1x1]
%onnx::Conv_813[FLOAT, 184x1x3x3]
%onnx::Conv_816[FLOAT, 184x184x1x1]
%onnx::Conv_819[FLOAT, 1104x184x1x1]
%onnx::Conv_822[FLOAT, 1104x1x3x3]
%onnx::Conv_825[FLOAT, 184x1104x1x1]
%onnx::Conv_828[FLOAT, 184x92x1x1]
%onnx::Conv_831[FLOAT, 184x1x5x5]
%onnx::Conv_834[FLOAT, 352x92x1x1]
%onnx::Conv_835[FLOAT, 352]
%onnx::Conv_837[FLOAT, 1504x352x1x1]
%onnx::Conv_838[FLOAT, 1504]
) {
%onnx::Conv_832 = Identity(%onnx::Conv_799)
%onnx::Conv_829 = Identity(%onnx::Conv_799)
%onnx::Conv_826 = Identity(%onnx::Conv_799)
%onnx::Conv_823 = Identity(%onnx::Conv_802)
%onnx::Conv_820 = Identity(%onnx::Conv_802)
%onnx::Conv_817 = Identity(%onnx::Conv_799)
%onnx::Conv_814 = Identity(%onnx::Conv_799)
%onnx::Conv_811 = Identity(%onnx::Conv_799)
%onnx::Conv_808 = Identity(%onnx::Conv_799)
%onnx::Conv_805 = Identity(%onnx::Conv_802)
%onnx::Conv_796 = Identity(%onnx::Conv_763)
%onnx::Conv_793 = Identity(%onnx::Conv_763)
%onnx::Conv_790 = Identity(%onnx::Conv_763)
%onnx::Conv_787 = Identity(%onnx::Conv_766)
%onnx::Conv_784 = Identity(%onnx::Conv_766)
%onnx::Conv_781 = Identity(%onnx::Conv_763)
%onnx::Conv_778 = Identity(%onnx::Conv_766)
%onnx::Conv_775 = Identity(%onnx::Conv_766)
%onnx::Conv_772 = Identity(%onnx::Conv_763)
%onnx::Conv_769 = Identity(%onnx::Conv_766)
%onnx::Conv_760 = Identity(%onnx::Conv_727)
%onnx::Conv_757 = Identity(%onnx::Conv_727)
%onnx::Conv_754 = Identity(%onnx::Conv_727)
%onnx::Conv_751 = Identity(%onnx::Conv_709)
%onnx::Conv_748 = Identity(%onnx::Conv_709)
%onnx::Conv_745 = Identity(%onnx::Conv_727)
%onnx::Conv_742 = Identity(%onnx::Conv_727)
%onnx::Conv_739 = Identity(%onnx::Conv_727)
%onnx::Conv_736 = Identity(%onnx::Conv_727)
%onnx::Conv_733 = Identity(%onnx::Conv_709)
%onnx::Conv_730 = Identity(%onnx::Conv_709)
%onnx::Conv_724 = Identity(%onnx::Conv_697)
%onnx::Conv_721 = Identity(%onnx::Conv_718)
%onnx::Conv_715 = Identity(%onnx::Conv_697)
%onnx::Conv_712 = Identity(%onnx::Conv_709)
%onnx::Conv_706 = Identity(%onnx::Conv_697)
%onnx::Conv_703 = Identity(%onnx::Conv_697)
%onnx::Conv_700 = Identity(%onnx::Conv_697)
%onnx::Conv_694 = Identity(%onnx::Conv_661)
%onnx::Conv_691 = Identity(%onnx::Conv_661)
%onnx::Conv_688 = Identity(%onnx::Conv_661)
%onnx::Conv_685 = Identity(%onnx::Conv_682)
%onnx::Conv_679 = Identity(%onnx::Conv_661)
%onnx::Conv_676 = Identity(%onnx::Conv_673)
%onnx::Conv_670 = Identity(%onnx::Conv_661)
%onnx::Conv_667 = Identity(%onnx::Conv_661)
%onnx::Conv_664 = Identity(%onnx::Conv_661)
%onnx::Conv_658 = Identity(%onnx::Conv_655)
%/stem/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%input.1, %onnx::Conv_651, %onnx::Conv_652)
%/stem/relu/Relu_output_0 = Relu(%/stem/conv/Conv_output_0)
%/cells.1/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/stem/relu/Relu_output_0, %onnx::Conv_654, %onnx::Conv_655)
%/cells.1/nl/Relu_output_0 = Relu(%/cells.1/conv1/Conv_output_0)
%/cells.1/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 48, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.1/nl/Relu_output_0, %onnx::Conv_657, %onnx::Conv_658)
%/cells.1/nl_1/Relu_output_0 = Relu(%/cells.1/conv2/Conv_output_0)
%/cells.1/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/nl_1/Relu_output_0, %onnx::Conv_660, %onnx::Conv_661)
%/cells.2/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.1/conv3/Conv_output_0, %onnx::Conv_663, %onnx::Conv_664)
%/cells.2/nl/Relu_output_0 = Relu(%/cells.2/conv1/Conv_output_0)
%/cells.2/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_output_0 = Reshape(%/cells.2/nl/Relu_output_0, %/cells.2/shuffle/Constant_output_0)
%/cells.2/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.2/shuffle/Reshape_output_0)
%/cells.2/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.2/shuffle/Reshape_1_output_0 = Reshape(%/cells.2/shuffle/Transpose_output_0, %/cells.2/shuffle/Constant_1_output_0)
%/cells.2/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.2/shuffle/Reshape_1_output_0, %onnx::Conv_666, %onnx::Conv_667)
%/cells.2/nl_1/Relu_output_0 = Relu(%/cells.2/conv2/Conv_output_0)
%/cells.2/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/nl_1/Relu_output_0, %onnx::Conv_669, %onnx::Conv_670)
%/cells.2/Add_output_0 = Add(%/cells.2/conv3/Conv_output_0, %/cells.1/conv3/Conv_output_0)
%/cells.3/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.2/Add_output_0, %onnx::Conv_672, %onnx::Conv_673)
%/cells.3/nl/Relu_output_0 = Relu(%/cells.3/conv1/Conv_output_0)
%/cells.3/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 144, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.3/nl/Relu_output_0, %onnx::Conv_675, %onnx::Conv_676)
%/cells.3/nl_1/Relu_output_0 = Relu(%/cells.3/conv2/Conv_output_0)
%/cells.3/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/nl_1/Relu_output_0, %onnx::Conv_678, %onnx::Conv_679)
%/cells.3/Add_output_0 = Add(%/cells.3/conv3/Conv_output_0, %/cells.2/Add_output_0)
%/cells.4/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.3/Add_output_0, %onnx::Conv_681, %onnx::Conv_682)
%/cells.4/nl/Relu_output_0 = Relu(%/cells.4/conv1/Conv_output_0)
%/cells.4/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 72, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.4/nl/Relu_output_0, %onnx::Conv_684, %onnx::Conv_685)
%/cells.4/nl_1/Relu_output_0 = Relu(%/cells.4/conv2/Conv_output_0)
%/cells.4/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/nl_1/Relu_output_0, %onnx::Conv_687, %onnx::Conv_688)
%/cells.4/Add_output_0 = Add(%/cells.4/conv3/Conv_output_0, %/cells.3/Add_output_0)
%/cells.5/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.4/Add_output_0, %onnx::Conv_690, %onnx::Conv_691)
%/cells.5/nl/Relu_output_0 = Relu(%/cells.5/conv1/Conv_output_0)
%/cells.5/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 24, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.5/nl/Relu_output_0, %onnx::Conv_693, %onnx::Conv_694)
%/cells.5/nl_1/Relu_output_0 = Relu(%/cells.5/conv2/Conv_output_0)
%/cells.5/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/nl_1/Relu_output_0, %onnx::Conv_696, %onnx::Conv_697)
%/cells.6/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.5/conv3/Conv_output_0, %onnx::Conv_699, %onnx::Conv_700)
%/cells.6/nl/Relu_output_0 = Relu(%/cells.6/conv1/Conv_output_0)
%/cells.6/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_output_0 = Reshape(%/cells.6/nl/Relu_output_0, %/cells.6/shuffle/Constant_output_0)
%/cells.6/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.6/shuffle/Reshape_output_0)
%/cells.6/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.6/shuffle/Reshape_1_output_0 = Reshape(%/cells.6/shuffle/Transpose_output_0, %/cells.6/shuffle/Constant_1_output_0)
%/cells.6/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 32, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.6/shuffle/Reshape_1_output_0, %onnx::Conv_702, %onnx::Conv_703)
%/cells.6/nl_1/Relu_output_0 = Relu(%/cells.6/conv2/Conv_output_0)
%/cells.6/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/nl_1/Relu_output_0, %onnx::Conv_705, %onnx::Conv_706)
%/cells.6/Add_output_0 = Add(%/cells.6/conv3/Conv_output_0, %/cells.5/conv3/Conv_output_0)
%/cells.7/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.6/Add_output_0, %onnx::Conv_708, %onnx::Conv_709)
%/cells.7/nl/Relu_output_0 = Relu(%/cells.7/conv1/Conv_output_0)
%/cells.7/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.7/nl/Relu_output_0, %onnx::Conv_711, %onnx::Conv_712)
%/cells.7/nl_1/Relu_output_0 = Relu(%/cells.7/conv2/Conv_output_0)
%/cells.7/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/nl_1/Relu_output_0, %onnx::Conv_714, %onnx::Conv_715)
%/cells.7/Add_output_0 = Add(%/cells.7/conv3/Conv_output_0, %/cells.6/Add_output_0)
%/cells.8/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.7/Add_output_0, %onnx::Conv_717, %onnx::Conv_718)
%/cells.8/nl/Relu_output_0 = Relu(%/cells.8/conv1/Conv_output_0)
%/cells.8/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 96, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.8/nl/Relu_output_0, %onnx::Conv_720, %onnx::Conv_721)
%/cells.8/nl_1/Relu_output_0 = Relu(%/cells.8/conv2/Conv_output_0)
%/cells.8/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.8/nl_1/Relu_output_0, %onnx::Conv_723, %onnx::Conv_724)
%/cells.8/Add_output_0 = Add(%/cells.8/conv3/Conv_output_0, %/cells.7/Add_output_0)
%/cells.9/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [2, 2]](%/cells.8/Add_output_0, %onnx::Conv_726, %onnx::Conv_727)
%/cells.9/relu/Relu_output_0 = Relu(%/cells.9/conv/Conv_output_0)
%/cells.10/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.9/relu/Relu_output_0, %onnx::Conv_729, %onnx::Conv_730)
%/cells.10/nl/Relu_output_0 = Relu(%/cells.10/conv1/Conv_output_0)
%/cells.10/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.10/nl/Relu_output_0, %onnx::Conv_732, %onnx::Conv_733)
%/cells.10/nl_1/Relu_output_0 = Relu(%/cells.10/conv2/Conv_output_0)
%/cells.10/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/nl_1/Relu_output_0, %onnx::Conv_735, %onnx::Conv_736)
%/cells.10/Add_output_0 = Add(%/cells.10/conv3/Conv_output_0, %/cells.9/relu/Relu_output_0)
%/cells.11/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.10/Add_output_0, %onnx::Conv_738, %onnx::Conv_739)
%/cells.11/nl/Relu_output_0 = Relu(%/cells.11/conv1/Conv_output_0)
%/cells.11/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_output_0 = Reshape(%/cells.11/nl/Relu_output_0, %/cells.11/shuffle/Constant_output_0)
%/cells.11/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.11/shuffle/Reshape_output_0)
%/cells.11/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.11/shuffle/Reshape_1_output_0 = Reshape(%/cells.11/shuffle/Transpose_output_0, %/cells.11/shuffle/Constant_1_output_0)
%/cells.11/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.11/shuffle/Reshape_1_output_0, %onnx::Conv_741, %onnx::Conv_742)
%/cells.11/nl_1/Relu_output_0 = Relu(%/cells.11/conv2/Conv_output_0)
%/cells.11/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/nl_1/Relu_output_0, %onnx::Conv_744, %onnx::Conv_745)
%/cells.11/Add_output_0 = Add(%/cells.11/conv3/Conv_output_0, %/cells.10/Add_output_0)
%/cells.12/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.11/Add_output_0, %onnx::Conv_747, %onnx::Conv_748)
%/cells.12/nl/Relu_output_0 = Relu(%/cells.12/conv1/Conv_output_0)
%/cells.12/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 192, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.12/nl/Relu_output_0, %onnx::Conv_750, %onnx::Conv_751)
%/cells.12/nl_1/Relu_output_0 = Relu(%/cells.12/conv2/Conv_output_0)
%/cells.12/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/nl_1/Relu_output_0, %onnx::Conv_753, %onnx::Conv_754)
%/cells.12/Add_output_0 = Add(%/cells.12/conv3/Conv_output_0, %/cells.11/Add_output_0)
%/cells.13/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.12/Add_output_0, %onnx::Conv_756, %onnx::Conv_757)
%/cells.13/nl/Relu_output_0 = Relu(%/cells.13/conv1/Conv_output_0)
%/cells.13/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 64, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.13/nl/Relu_output_0, %onnx::Conv_759, %onnx::Conv_760)
%/cells.13/nl_1/Relu_output_0 = Relu(%/cells.13/conv2/Conv_output_0)
%/cells.13/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/nl_1/Relu_output_0, %onnx::Conv_762, %onnx::Conv_763)
%/cells.14/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.13/conv3/Conv_output_0, %onnx::Conv_765, %onnx::Conv_766)
%/cells.14/nl/Relu_output_0 = Relu(%/cells.14/conv1/Conv_output_0)
%/cells.14/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.14/nl/Relu_output_0, %onnx::Conv_768, %onnx::Conv_769)
%/cells.14/nl_1/Relu_output_0 = Relu(%/cells.14/conv2/Conv_output_0)
%/cells.14/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/nl_1/Relu_output_0, %onnx::Conv_771, %onnx::Conv_772)
%/cells.14/Add_output_0 = Add(%/cells.14/conv3/Conv_output_0, %/cells.13/conv3/Conv_output_0)
%/cells.15/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.14/Add_output_0, %onnx::Conv_774, %onnx::Conv_775)
%/cells.15/nl/Relu_output_0 = Relu(%/cells.15/conv1/Conv_output_0)
%/cells.15/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.15/nl/Relu_output_0, %onnx::Conv_777, %onnx::Conv_778)
%/cells.15/nl_1/Relu_output_0 = Relu(%/cells.15/conv2/Conv_output_0)
%/cells.15/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/nl_1/Relu_output_0, %onnx::Conv_780, %onnx::Conv_781)
%/cells.15/Add_output_0 = Add(%/cells.15/conv3/Conv_output_0, %/cells.14/Add_output_0)
%/cells.16/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.15/Add_output_0, %onnx::Conv_783, %onnx::Conv_784)
%/cells.16/nl/Relu_output_0 = Relu(%/cells.16/conv1/Conv_output_0)
%/cells.16/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 672, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.16/nl/Relu_output_0, %onnx::Conv_786, %onnx::Conv_787)
%/cells.16/nl_1/Relu_output_0 = Relu(%/cells.16/conv2/Conv_output_0)
%/cells.16/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/nl_1/Relu_output_0, %onnx::Conv_789, %onnx::Conv_790)
%/cells.16/Add_output_0 = Add(%/cells.16/conv3/Conv_output_0, %/cells.15/Add_output_0)
%/cells.17/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.16/Add_output_0, %onnx::Conv_792, %onnx::Conv_793)
%/cells.17/nl/Relu_output_0 = Relu(%/cells.17/conv1/Conv_output_0)
%/cells.17/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 112, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [2, 2]](%/cells.17/nl/Relu_output_0, %onnx::Conv_795, %onnx::Conv_796)
%/cells.17/nl_1/Relu_output_0 = Relu(%/cells.17/conv2/Conv_output_0)
%/cells.17/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/nl_1/Relu_output_0, %onnx::Conv_798, %onnx::Conv_799)
%/cells.18/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.17/conv3/Conv_output_0, %onnx::Conv_801, %onnx::Conv_802)
%/cells.18/nl/Relu_output_0 = Relu(%/cells.18/conv1/Conv_output_0)
%/cells.18/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.18/nl/Relu_output_0, %onnx::Conv_804, %onnx::Conv_805)
%/cells.18/nl_1/Relu_output_0 = Relu(%/cells.18/conv2/Conv_output_0)
%/cells.18/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/nl_1/Relu_output_0, %onnx::Conv_807, %onnx::Conv_808)
%/cells.18/Add_output_0 = Add(%/cells.18/conv3/Conv_output_0, %/cells.17/conv3/Conv_output_0)
%/cells.19/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.18/Add_output_0, %onnx::Conv_810, %onnx::Conv_811)
%/cells.19/nl/Relu_output_0 = Relu(%/cells.19/conv1/Conv_output_0)
%/cells.19/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.19/nl/Relu_output_0, %onnx::Conv_813, %onnx::Conv_814)
%/cells.19/nl_1/Relu_output_0 = Relu(%/cells.19/conv2/Conv_output_0)
%/cells.19/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/nl_1/Relu_output_0, %onnx::Conv_816, %onnx::Conv_817)
%/cells.19/Add_output_0 = Add(%/cells.19/conv3/Conv_output_0, %/cells.18/Add_output_0)
%/cells.20/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.19/Add_output_0, %onnx::Conv_819, %onnx::Conv_820)
%/cells.20/nl/Relu_output_0 = Relu(%/cells.20/conv1/Conv_output_0)
%/cells.20/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 1104, kernel_shape = [3, 3], pads = [1, 1, 1, 1], strides = [1, 1]](%/cells.20/nl/Relu_output_0, %onnx::Conv_822, %onnx::Conv_823)
%/cells.20/nl_1/Relu_output_0 = Relu(%/cells.20/conv2/Conv_output_0)
%/cells.20/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/nl_1/Relu_output_0, %onnx::Conv_825, %onnx::Conv_826)
%/cells.20/Add_output_0 = Add(%/cells.20/conv3/Conv_output_0, %/cells.19/Add_output_0)
%/cells.21/conv1/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.20/Add_output_0, %onnx::Conv_828, %onnx::Conv_829)
%/cells.21/nl/Relu_output_0 = Relu(%/cells.21/conv1/Conv_output_0)
%/cells.21/shuffle/Constant_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_output_0 = Reshape(%/cells.21/nl/Relu_output_0, %/cells.21/shuffle/Constant_output_0)
%/cells.21/shuffle/Transpose_output_0 = Transpose[perm = [0, 2, 1, 3, 4]](%/cells.21/shuffle/Reshape_output_0)
%/cells.21/shuffle/Constant_1_output_0 = Constant[value = <Tensor>]()
%/cells.21/shuffle/Reshape_1_output_0 = Reshape(%/cells.21/shuffle/Transpose_output_0, %/cells.21/shuffle/Constant_1_output_0)
%/cells.21/conv2/Conv_output_0 = Conv[dilations = [1, 1], group = 184, kernel_shape = [5, 5], pads = [2, 2, 2, 2], strides = [1, 1]](%/cells.21/shuffle/Reshape_1_output_0, %onnx::Conv_831, %onnx::Conv_832)
%/cells.21/nl_1/Relu_output_0 = Relu(%/cells.21/conv2/Conv_output_0)
%/cells.21/conv3/Conv_output_0 = Conv[dilations = [1, 1], group = 2, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/nl_1/Relu_output_0, %onnx::Conv_834, %onnx::Conv_835)
%/header/conv/Conv_output_0 = Conv[dilations = [1, 1], group = 1, kernel_shape = [1, 1], pads = [0, 0, 0, 0], strides = [1, 1]](%/cells.21/conv3/Conv_output_0, %onnx::Conv_837, %onnx::Conv_838)
%/header/relu/Relu_output_0 = Relu(%/header/conv/Conv_output_0)
%/avgpool/GlobalAveragePool_output_0 = GlobalAveragePool(%/header/relu/Relu_output_0)
%/Constant_output_0 = Constant[value = <Tensor>]()
%/Reshape_output_0 = Reshape(%/avgpool/GlobalAveragePool_output_0, %/Constant_output_0)
%649 = Gemm[alpha = 1, beta = 1, transB = 1](%/Reshape_output_0, %fc.weight, %fc.bias)
return %649
}
|
val_accuracy
| 0
| 92,436,352
| 2,327,244
|
{'zcp_synflow': 80.93466393646146, 'zcp_zen': 72.44190216064453, 'zcp_epe_nas': 8.997486308822443, 'zcp_fisher': 0.08677017688751221, 'zcp_flops': 92436352.0, 'zcp_grad_norm': 24.373563766479492, 'zcp_grasp': -0.09206008911132812, 'zcp_jacov': -16.05456913315527, 'zcp_l2_norm': 691.0089721679688, 'zcp_nwot': 215.14171178968377, 'zcp_params': 2327244.0, 'zcp_plain': -0.0014797864714637399, 'zcp_snip': 42.238372802734375, 'lat_1080ti_1': 0.5375788028005415, 'lat_1080ti_32': 0.5997814693215419, 'lat_1080ti_64': 0.5993228525393325, 'lat_2080ti_1': 0.6184901266719578, 'lat_2080ti_32': 0.6015095287269183, 'lat_2080ti_64': 0.5972629873604445, 'lat_essential_ph_1': 0.5094339622641509, 'lat_eyeriss': 0.6909302624099171, 'lat_fpga': 0.7896377363718223, 'lat_gold_6226': 0.5431141791963802, 'lat_gold_6240': 0.768274791470988, 'lat_pixel2': 0.5434782608695652, 'lat_pixel3': 0.7292644216014134, 'lat_raspi4': 0.7262841140602152, 'lat_samsung_a50': 0.37894736842105264, 'lat_samsung_s7': 0.25196850393700787, 'lat_silver_4114': 0.8123406989775338, 'lat_silver_4210r': 0.7699441438407684, 'lat_titan_rtx_1': 0.595339892808536, 'lat_titan_rtx_32': 0.5781422947640805, 'lat_titan_rtx_64': 0.5985195897721384, 'lat_titanx_1': 0.32879142181290927, 'lat_titanx_32': 0.6778742153189106, 'lat_titanx_64': 0.5641302604595094, 'lat_titanxp_1': 0.5740171177950422, 'lat_titanxp_32': 0.6147581826796062, 'lat_titanxp_64': 0.598972838936813}
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.