| 38 | return out |
| 39 | |
| 40 | class Bottleneck(nn.Module): |
| 41 | expansion = 4 |
| 42 | |
| 43 | def __init__(self, in_channel, out_channel, stride=1, downsample=None, groups=1, width_per_group=64): |
| 44 | super(Bottleneck, self).__init__() |
| 45 | |
| 46 | width = int(out_channel * (width_per_group / 64.)) * groups |
| 47 | |
| 48 | self.conv1 = nn.Conv2d(in_channels=in_channel, out_channels=width, |
| 49 | kernel_size=1, stride=1, bias=False) |
| 50 | |
| 51 | self.bn1 = nn.BatchNorm2d(width) |
| 52 | |
| 53 | self.conv2 = nn.Conv2d(in_channels=width, out_channels=width, groups=groups, |
| 54 | kernel_size=3, stride=stride, bias=False, padding=1) |
| 55 | |
| 56 | self.bn2 = nn.BatchNorm2d(width) |
| 57 | |
| 58 | self.conv3 = nn.Conv2d(in_channels=width, out_channels=out_channel*self.expansion, |
| 59 | kernel_size=1, stride=1, bias=False) |
| 60 | self.bn3 = nn.BatchNorm2d(out_channel*self.expansion) |
| 61 | self.relu = nn.ReLU(inplace=True) |
| 62 | self.downsample = downsample |
| 63 | |
| 64 | def forward(self, x): |
| 65 | identity = x |
| 66 | if self.downsample is not None: |
| 67 | identity = self.downsample(x) |
| 68 | |
| 69 | out = self.conv1(x) |
| 70 | out = self.bn1(out) |
| 71 | out = self.relu(out) |
| 72 | |
| 73 | out = self.conv2(out) |
| 74 | out = self.bn2(out) |
| 75 | out = self.relu(out) |
| 76 | |
| 77 | out = self.conv3(out) |
| 78 | out = self.bn3(out) |
| 79 | |
| 80 | out += identity |
| 81 | out = self.relu(out) |
| 82 | |
| 83 | return out |
| 84 | |
| 85 | class ResNet(nn.Module): |
| 86 | def __init__(self, |
nothing calls this directly
no outgoing calls
no test coverage detected