Inceptionv4 Kernel padding mismatch

vision
1

hi i get below error
when i make changes to inceptionv4
Calculated padded input size per channel: (2 x 2). Kernel size: (3 x 3). Kernel size can’t be greater than actual input size at /opt/conda/conda-bld/pytorch-nightly_1540121100527/work/aten/src/THNN/generic/SpatialConvolutionMM.c:50
Here is how i customize the model

def dense(pre=True):

#model = models.densenet121(pretrained=pre)
model = m1.__dict__['inceptionv4'](pretrained='imagenet')
w=model.features[0].conv.weight
model.features[0].conv=nn.Conv2d(4, 32, kernel_size=3, stride=2,padding=0, bias=False)
#model.classifier = (nn.Linear(1024, 28))
model.last_linear= nn.Linear(in_features=1536, out_features=28, bias=True)

print(w.shape)
model.features[0].conv.weight=torch.nn.Parameter(torch.cat((w, w[:,:1,:,:]),dim=1))
print(model.features[0].conv.weight.shape)
return model

Below is the model:

Strange thing i dont get this issue when i run my kernel in Kaggle.
weights size
torch.Size([32, 3, 3, 3])
torch.Size([32, 4, 3, 3])

InceptionV4(
(features): Sequential(
(0): BasicConv2d(
(conv): Conv2d(4, 32, kernel_size=(3, 3), stride=(2, 2), bias=False)
(bn): BatchNorm2d(32, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), bias=False)
(bn): BatchNorm2d(32, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(32, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn): BatchNorm2d(64, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(3): Mixed_3a(
(maxpool): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=False)
(conv): BasicConv2d(
(conv): Conv2d(64, 96, kernel_size=(3, 3), stride=(2, 2), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(4): Mixed_4a(
(branch0): Sequential(
(0): BasicConv2d(
(conv): Conv2d(160, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(64, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(64, 96, kernel_size=(3, 3), stride=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch1): Sequential(
(0): BasicConv2d(
(conv): Conv2d(160, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(64, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(64, 64, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(64, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(64, 64, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(64, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(3): BasicConv2d(
(conv): Conv2d(64, 96, kernel_size=(3, 3), stride=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
)
(5): Mixed_5a(
(conv): BasicConv2d(
(conv): Conv2d(192, 192, kernel_size=(3, 3), stride=(2, 2), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(maxpool): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=False)
)
(6): Inception_A(
(branch0): BasicConv2d(
(conv): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1): Sequential(
(0): BasicConv2d(
(conv): Conv2d(384, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(64, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(64, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch2): Sequential(
(0): BasicConv2d(
(conv): Conv2d(384, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(64, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(64, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch3): Sequential(
(0): AvgPool2d(kernel_size=3, stride=1, padding=1)
(1): BasicConv2d(
(conv): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
)
(7): Inception_A(
(branch0): BasicConv2d(
(conv): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1): Sequential(
(0): BasicConv2d(
(conv): Conv2d(384, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(64, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(64, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch2): Sequential(
(0): BasicConv2d(
(conv): Conv2d(384, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(64, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(64, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch3): Sequential(
(0): AvgPool2d(kernel_size=3, stride=1, padding=1)
(1): BasicConv2d(
(conv): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
)
(8): Inception_A(
(branch0): BasicConv2d(
(conv): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1): Sequential(
(0): BasicConv2d(
(conv): Conv2d(384, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(64, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(64, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch2): Sequential(
(0): BasicConv2d(
(conv): Conv2d(384, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(64, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(64, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch3): Sequential(
(0): AvgPool2d(kernel_size=3, stride=1, padding=1)
(1): BasicConv2d(
(conv): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
)
(9): Inception_A(
(branch0): BasicConv2d(
(conv): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1): Sequential(
(0): BasicConv2d(
(conv): Conv2d(384, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(64, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(64, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch2): Sequential(
(0): BasicConv2d(
(conv): Conv2d(384, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(64, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(64, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(96, 96, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch3): Sequential(
(0): AvgPool2d(kernel_size=3, stride=1, padding=1)
(1): BasicConv2d(
(conv): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(96, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
)

2

continue:

(10): Reduction_A(
(branch0): BasicConv2d(
(conv): Conv2d(384, 384, kernel_size=(3, 3), stride=(2, 2), bias=False)
(bn): BatchNorm2d(384, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1): Sequential(
(0): BasicConv2d(
(conv): Conv2d(384, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 224, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(224, 256, kernel_size=(3, 3), stride=(2, 2), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch2): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=False)
)
(11): Inception_B(
(branch0): BasicConv2d(
(conv): Conv2d(1024, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(384, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 224, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(224, 256, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch2): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 192, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(192, 224, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(3): BasicConv2d(
(conv): Conv2d(224, 224, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(4): BasicConv2d(
(conv): Conv2d(224, 256, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch3): Sequential(
(0): AvgPool2d(kernel_size=3, stride=1, padding=1)
(1): BasicConv2d(
(conv): Conv2d(1024, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(128, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
)
(12): Inception_B(
(branch0): BasicConv2d(
(conv): Conv2d(1024, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(384, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 224, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(224, 256, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch2): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 192, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(192, 224, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(3): BasicConv2d(
(conv): Conv2d(224, 224, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(4): BasicConv2d(
(conv): Conv2d(224, 256, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch3): Sequential(
(0): AvgPool2d(kernel_size=3, stride=1, padding=1)
(1): BasicConv2d(
(conv): Conv2d(1024, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(128, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
)
(13): Inception_B(
(branch0): BasicConv2d(
(conv): Conv2d(1024, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(384, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 224, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(224, 256, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch2): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 192, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(192, 224, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(3): BasicConv2d(
(conv): Conv2d(224, 224, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(4): BasicConv2d(
(conv): Conv2d(224, 256, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch3): Sequential(
(0): AvgPool2d(kernel_size=3, stride=1, padding=1)
(1): BasicConv2d(
(conv): Conv2d(1024, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(128, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
)
(14): Inception_B(
(branch0): BasicConv2d(
(conv): Conv2d(1024, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(384, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 224, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(224, 256, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch2): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 192, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(192, 224, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(3): BasicConv2d(
(conv): Conv2d(224, 224, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(4): BasicConv2d(
(conv): Conv2d(224, 256, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch3): Sequential(
(0): AvgPool2d(kernel_size=3, stride=1, padding=1)
(1): BasicConv2d(
(conv): Conv2d(1024, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(128, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
)
(15): Inception_B(
(branch0): BasicConv2d(
(conv): Conv2d(1024, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(384, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 224, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(224, 256, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch2): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 192, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(192, 224, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(3): BasicConv2d(
(conv): Conv2d(224, 224, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(4): BasicConv2d(
(conv): Conv2d(224, 256, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch3): Sequential(
(0): AvgPool2d(kernel_size=3, stride=1, padding=1)
(1): BasicConv2d(
(conv): Conv2d(1024, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(128, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
)
(16): Inception_B(
(branch0): BasicConv2d(
(conv): Conv2d(1024, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(384, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 224, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(224, 256, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch2): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 192, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(192, 224, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(3): BasicConv2d(
(conv): Conv2d(224, 224, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(4): BasicConv2d(
(conv): Conv2d(224, 256, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch3): Sequential(
(0): AvgPool2d(kernel_size=3, stride=1, padding=1)
(1): BasicConv2d(
(conv): Conv2d(1024, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(128, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
)
(17): Inception_B(
(branch0): BasicConv2d(
(conv): Conv2d(1024, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(384, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 224, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(224, 256, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch2): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 192, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(192, 224, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(3): BasicConv2d(
(conv): Conv2d(224, 224, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(224, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(4): BasicConv2d(
(conv): Conv2d(224, 256, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch3): Sequential(
(0): AvgPool2d(kernel_size=3, stride=1, padding=1)
(1): BasicConv2d(
(conv): Conv2d(1024, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(128, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
)
(18): Reduction_B(
(branch0): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 192, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(192, 192, kernel_size=(3, 3), stride=(2, 2), bias=False)
(bn): BatchNorm2d(192, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch1): Sequential(
(0): BasicConv2d(
(conv): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(1): BasicConv2d(
(conv): Conv2d(256, 256, kernel_size=(1, 7), stride=(1, 1), padding=(0, 3), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(2): BasicConv2d(
(conv): Conv2d(256, 320, kernel_size=(7, 1), stride=(1, 1), padding=(3, 0), bias=False)
(bn): BatchNorm2d(320, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(3): BasicConv2d(
(conv): Conv2d(320, 320, kernel_size=(3, 3), stride=(2, 2), bias=False)
(bn): BatchNorm2d(320, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
(branch2): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=False)
)
(19): Inception_C(
(branch0): BasicConv2d(
(conv): Conv2d(1536, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1_0): BasicConv2d(
(conv): Conv2d(1536, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(384, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1_1a): BasicConv2d(
(conv): Conv2d(384, 256, kernel_size=(1, 3), stride=(1, 1), padding=(0, 1), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1_1b): BasicConv2d(
(conv): Conv2d(384, 256, kernel_size=(3, 1), stride=(1, 1), padding=(1, 0), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch2_0): BasicConv2d(
(conv): Conv2d(1536, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(384, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch2_1): BasicConv2d(
(conv): Conv2d(384, 448, kernel_size=(3, 1), stride=(1, 1), padding=(1, 0), bias=False)
(bn): BatchNorm2d(448, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch2_2): BasicConv2d(
(conv): Conv2d(448, 512, kernel_size=(1, 3), stride=(1, 1), padding=(0, 1), bias=False)
(bn): BatchNorm2d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch2_3a): BasicConv2d(
(conv): Conv2d(512, 256, kernel_size=(1, 3), stride=(1, 1), padding=(0, 1), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch2_3b): BasicConv2d(
(conv): Conv2d(512, 256, kernel_size=(3, 1), stride=(1, 1), padding=(1, 0), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch3): Sequential(
(0): AvgPool2d(kernel_size=3, stride=1, padding=1)
(1): BasicConv2d(
(conv): Conv2d(1536, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
)
(20): Inception_C(
(branch0): BasicConv2d(
(conv): Conv2d(1536, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1_0): BasicConv2d(
(conv): Conv2d(1536, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(384, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1_1a): BasicConv2d(
(conv): Conv2d(384, 256, kernel_size=(1, 3), stride=(1, 1), padding=(0, 1), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1_1b): BasicConv2d(
(conv): Conv2d(384, 256, kernel_size=(3, 1), stride=(1, 1), padding=(1, 0), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch2_0): BasicConv2d(
(conv): Conv2d(1536, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(384, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch2_1): BasicConv2d(
(conv): Conv2d(384, 448, kernel_size=(3, 1), stride=(1, 1), padding=(1, 0), bias=False)
(bn): BatchNorm2d(448, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch2_2): BasicConv2d(
(conv): Conv2d(448, 512, kernel_size=(1, 3), stride=(1, 1), padding=(0, 1), bias=False)
(bn): BatchNorm2d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch2_3a): BasicConv2d(
(conv): Conv2d(512, 256, kernel_size=(1, 3), stride=(1, 1), padding=(0, 1), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch2_3b): BasicConv2d(
(conv): Conv2d(512, 256, kernel_size=(3, 1), stride=(1, 1), padding=(1, 0), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch3): Sequential(
(0): AvgPool2d(kernel_size=3, stride=1, padding=1)
(1): BasicConv2d(
(conv): Conv2d(1536, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
)
(21): Inception_C(
(branch0): BasicConv2d(
(conv): Conv2d(1536, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1_0): BasicConv2d(
(conv): Conv2d(1536, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(384, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1_1a): BasicConv2d(
(conv): Conv2d(384, 256, kernel_size=(1, 3), stride=(1, 1), padding=(0, 1), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch1_1b): BasicConv2d(
(conv): Conv2d(384, 256, kernel_size=(3, 1), stride=(1, 1), padding=(1, 0), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch2_0): BasicConv2d(
(conv): Conv2d(1536, 384, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(384, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch2_1): BasicConv2d(
(conv): Conv2d(384, 448, kernel_size=(3, 1), stride=(1, 1), padding=(1, 0), bias=False)
(bn): BatchNorm2d(448, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch2_2): BasicConv2d(
(conv): Conv2d(448, 512, kernel_size=(1, 3), stride=(1, 1), padding=(0, 1), bias=False)
(bn): BatchNorm2d(512, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch2_3a): BasicConv2d(
(conv): Conv2d(512, 256, kernel_size=(1, 3), stride=(1, 1), padding=(0, 1), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch2_3b): BasicConv2d(
(conv): Conv2d(512, 256, kernel_size=(3, 1), stride=(1, 1), padding=(1, 0), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
(branch3): Sequential(
(0): AvgPool2d(kernel_size=3, stride=1, padding=1)
(1): BasicConv2d(
(conv): Conv2d(1536, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn): BatchNorm2d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
(relu): ReLU(inplace)
)
)
)
)
(avg_pool): AvgPool2d(kernel_size=8, stride=8, padding=0)
(last_linear): Linear(in_features=1536, out_features=28, bias=True)
)