Fixing AlexNet references (#5962)

* Fixing AlexNet references

* Fix doc rendering

docs/source/models/alexnet.rst

@@ -3,10 +3,12 @@ AlexNet
 
 .. currentmodule:: torchvision.models
 
-The AlexNet model is based on the
+The AlexNet model was originally introduced in the
 `ImageNet Classification with Deep Convolutional Neural Networks
-<https://papers.nips.cc/paper/2012/hash/c399862d3b9d6b76c8436e924a68c45b-Abstract.html>`__.
-paper.
+<https://papers.nips.cc/paper/2012/hash/c399862d3b9d6b76c8436e924a68c45b-Abstract.html>`__
+paper. The implemented architecture is slightly different from the original one,
+and is based on `One weird trick for parallelizing convolutional neural networks
+<https://arxiv.org/abs/1404.5997>`__.
 
 
 Model builders

torchvision/models/alexnet.py

@@ -72,11 +72,17 @@ class AlexNet_Weights(WeightsEnum):
 
 @handle_legacy_interface(weights=("pretrained", AlexNet_Weights.IMAGENET1K_V1))
 def alexnet(*, weights: Optional[AlexNet_Weights] = None, progress: bool = True, **kwargs: Any) -> AlexNet:
-    """AlexNet model architecture from the `ImageNet Classification with Deep Convolutional Neural Networks
-    <https://papers.nips.cc/paper/2012/hash/c399862d3b9d6b76c8436e924a68c45b-Abstract.html>`__ paper.
+    """AlexNet model architecture from `One weird trick for parallelizing convolutional neural networks <https://arxiv.org/abs/1404.5997>`__.
 
     The required minimum input size of the model is 63x63.
 
+    .. note::
+        AlexNet was originally introduced in the `ImageNet Classification with
+        Deep Convolutional Neural Networks
+        <https://papers.nips.cc/paper/2012/hash/c399862d3b9d6b76c8436e924a68c45b-Abstract.html>`__
+        paper. Our implementation is based instead on the "One weird trick"
+        paper above.
+
     Args:
         weights (:class:`~torchvision.models.AlexNet_Weights`, optional): The
             pretrained weights to use. See