The implementation of ResNet is different from official implementation in Caffe

The downsample part in each block/layer (not the skip connection part), the PyTorch do it in conv3x3 using stride=2, but official caffe version in conv1x1 with stride=2

conv1x1 -> caffe do it in here
conv3x3 -> pytorch do it in here
conv1x1

Here in Bottleneck:

        self.conv2 = nn.Conv2d(planes, planes, kernel_size=3, stride=stride,
                               padding=1, bias=False)

  (layer2): Sequential (
    (0): Bottleneck (
      (conv1): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
      (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True)
      (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)
       ...

but in caffe

layer {
	bottom: "res2c"
	top: "res3a_branch2a"
	name: "res3a_branch2a"
	type: "Convolution"
	convolution_param {
		num_output: 128
		kernel_size: 1
		pad: 0
		stride: 2
		bias_term: false
	}
}

Start at the Bottleneck implementation and inspect the shown conv2 stride placement. Compare it with the linked Caffe ResNet-101 prototxt, especially res3a_branch2a, then determine whether the implementation should match the official stride placement and verify the affected ResNet behavior.