Update ResNet-50 accuracy with Repeated Augmentation (#5201)

* Update ResNet-50 accuracy with Repeated Augmentations * Update the ResNet50 weights URL. Co-authored-by: Vasilis Vryniotis <[email protected]>
pytorch · Jan 17, 2022 · f923aeb · f923aeb
1 parent 09c5ddd
commit f923aeb
Showing 1 changed file with 4 additions and 4 deletions.
diff --git a/torchvision/prototype/models/resnet.py b/torchvision/prototype/models/resnet.py
@@ -108,16 +108,16 @@ class ResNet50_Weights(WeightsEnum):
         },
     )
     ImageNet1K_V2 = Weights(
-        url="https://download.pytorch.org/models/resnet50-f46c3f97.pth",
+        url="https://download.pytorch.org/models/resnet50-11ad3fa6.pth",
         transforms=partial(ImageNetEval, crop_size=224, resize_size=232),
         meta={
             **_COMMON_META,
             "architecture": "ResNet",
             "publication_year": 2015,
             "num_params": 25557032,
-            "recipe": "https://github.com/pytorch/vision/issues/3995#new-recipe-with-fixres",
-            "acc@1": 80.674,
-            "acc@5": 95.166,
+            "recipe": "https://github.com/pytorch/vision/issues/3995#issuecomment-1013906621",
+            "acc@1": 80.858,
+            "acc@5": 95.434,
         },
     )
     default = ImageNet1K_V2