release code

Dassl.ProGrad.pytorch/dassl/modeling/backbone/alexnet.py

@@ -0,0 +1,64 @@
+import torch
+import torch.nn as nn
+import torch.utils.model_zoo as model_zoo
+
+from .build import BACKBONE_REGISTRY
+from .backbone import Backbone
+
+model_urls = {
+    "alexnet": "https://download.pytorch.org/models/alexnet-owt-4df8aa71.pth",
+}
+
+
+class AlexNet(Backbone):
+
+    def __init__(self):
+        super().__init__()
+        self.features = nn.Sequential(
+            nn.Conv2d(3, 64, kernel_size=11, stride=4, padding=2),
+            nn.ReLU(inplace=True),
+            nn.MaxPool2d(kernel_size=3, stride=2),
+            nn.Conv2d(64, 192, kernel_size=5, padding=2),
+            nn.ReLU(inplace=True),
+            nn.MaxPool2d(kernel_size=3, stride=2),
+            nn.Conv2d(192, 384, kernel_size=3, padding=1),
+            nn.ReLU(inplace=True),
+            nn.Conv2d(384, 256, kernel_size=3, padding=1),
+            nn.ReLU(inplace=True),
+            nn.Conv2d(256, 256, kernel_size=3, padding=1),
+            nn.ReLU(inplace=True),
+            nn.MaxPool2d(kernel_size=3, stride=2),
+        )
+        self.avgpool = nn.AdaptiveAvgPool2d((6, 6))
+        # Note that self.classifier outputs features rather than logits
+        self.classifier = nn.Sequential(
+            nn.Dropout(),
+            nn.Linear(256 * 6 * 6, 4096),
+            nn.ReLU(inplace=True),
+            nn.Dropout(),
+            nn.Linear(4096, 4096),
+            nn.ReLU(inplace=True),
+        )
+
+        self._out_features = 4096
+
+    def forward(self, x):
+        x = self.features(x)
+        x = self.avgpool(x)
+        x = torch.flatten(x, 1)
+        return self.classifier(x)
+
+
+def init_pretrained_weights(model, model_url):
+    pretrain_dict = model_zoo.load_url(model_url)
+    model.load_state_dict(pretrain_dict, strict=False)
+
+
+@BACKBONE_REGISTRY.register()
+def alexnet(pretrained=True, **kwargs):
+    model = AlexNet()
+
+    if pretrained:
+        init_pretrained_weights(model, model_urls["alexnet"])
+
+    return model