Spaces:
Sleeping
Sleeping
| import torch | |
| import torch.nn as nn | |
| import torch.nn.functional as F | |
| class Bottleneck(nn.Module): | |
| """Bottleneck residual block for ResNet-50/101/152""" | |
| expansion = 4 | |
| def __init__(self, in_channels, out_channels, stride=1, downsample=None, dropout=0.0): | |
| super(Bottleneck, self).__init__() | |
| self.conv1 = nn.Conv2d(in_channels, out_channels, kernel_size=1, bias=False) | |
| self.bn1 = nn.BatchNorm2d(out_channels) | |
| self.conv2 = nn.Conv2d(out_channels, out_channels, kernel_size=3, stride=stride, padding=1, bias=False) | |
| self.bn2 = nn.BatchNorm2d(out_channels) | |
| self.conv3 = nn.Conv2d(out_channels, out_channels * self.expansion, kernel_size=1, bias=False) | |
| self.bn3 = nn.BatchNorm2d(out_channels * self.expansion) | |
| self.downsample = downsample | |
| self.dropout = nn.Dropout2d(dropout) if dropout > 0 else None | |
| def forward(self, x): | |
| identity = x | |
| out = self.conv1(x) | |
| out = self.bn1(out) | |
| out = F.relu(out) | |
| out = self.conv2(out) | |
| out = self.bn2(out) | |
| out = F.relu(out) | |
| if self.dropout is not None: | |
| out = self.dropout(out) | |
| out = self.conv3(out) | |
| out = self.bn3(out) | |
| if self.downsample is not None: | |
| identity = self.downsample(x) | |
| out += identity | |
| out = F.relu(out) | |
| return out | |
| class ResNet50(nn.Module): | |
| """ResNet-50 model for ImageNet-1K dataset""" | |
| def __init__(self, num_classes=1000, dropout=0.0): | |
| super(ResNet50, self).__init__() | |
| self.in_channels = 64 | |
| # Initial convolution layer for ImageNet (224x224 input) | |
| self.conv1 = nn.Conv2d(3, 64, kernel_size=7, stride=2, padding=3, bias=False) | |
| self.bn1 = nn.BatchNorm2d(64) | |
| self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1) | |
| # ResNet-50 architecture: [3, 4, 6, 3] blocks per layer group | |
| self.layer1 = self._make_layer(64, 3, stride=1, dropout=dropout) | |
| self.layer2 = self._make_layer(128, 4, stride=2, dropout=dropout) | |
| self.layer3 = self._make_layer(256, 6, stride=2, dropout=dropout) | |
| self.layer4 = self._make_layer(512, 3, stride=2, dropout=dropout) | |
| # Final layers for ImageNet | |
| self.avgpool = nn.AdaptiveAvgPool2d((1, 1)) | |
| self.dropout = nn.Dropout(0.5) # Standard dropout for ImageNet | |
| self.fc = nn.Linear(512 * Bottleneck.expansion, num_classes) | |
| # Initialize weights | |
| self._initialize_weights() | |
| def _make_layer(self, out_channels, blocks, stride, dropout=0.0): | |
| """Create a residual layer with specified number of blocks""" | |
| downsample = None | |
| if stride != 1 or self.in_channels != out_channels * Bottleneck.expansion: | |
| downsample = nn.Sequential( | |
| nn.Conv2d(self.in_channels, out_channels * Bottleneck.expansion, | |
| kernel_size=1, stride=stride, bias=False), | |
| nn.BatchNorm2d(out_channels * Bottleneck.expansion), | |
| ) | |
| layers = [] | |
| layers.append(Bottleneck(self.in_channels, out_channels, stride, downsample, dropout)) | |
| self.in_channels = out_channels * Bottleneck.expansion | |
| for _ in range(1, blocks): | |
| layers.append(Bottleneck(self.in_channels, out_channels, dropout=dropout)) | |
| return nn.Sequential(*layers) | |
| def _initialize_weights(self): | |
| """Initialize weights using He initialization""" | |
| for m in self.modules(): | |
| if isinstance(m, nn.Conv2d): | |
| nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu') | |
| elif isinstance(m, nn.BatchNorm2d): | |
| nn.init.constant_(m.weight, 1) | |
| nn.init.constant_(m.bias, 0) | |
| def forward(self, x): | |
| # Initial layers | |
| x = self.conv1(x) | |
| x = self.bn1(x) | |
| x = F.relu(x) | |
| x = self.maxpool(x) | |
| # Residual layers | |
| x = self.layer1(x) | |
| x = self.layer2(x) | |
| x = self.layer3(x) | |
| x = self.layer4(x) | |
| # Final layers | |
| x = self.avgpool(x) | |
| x = torch.flatten(x, 1) | |
| x = self.dropout(x) | |
| x = self.fc(x) | |
| return x | |
| if __name__ == "__main__": | |
| # Test the model | |
| model = ResNet50(num_classes=1000) | |
| print(f"Model parameters: {sum(p.numel() for p in model.parameters()):,}") | |
| print(f"Model trainable parameters: {sum(p.numel() for p in model.parameters() if p.requires_grad):,}") | |
| # Test forward pass | |
| x = torch.randn(1, 3, 224, 224) # ImageNet size (224x224) | |
| y = model(x) | |
| print(f"Input shape: {x.shape}") | |
| print(f"Output shape: {y.shape}") | |
| print(f"Expected output classes: {y.shape[1]}") | |