Vision: doc-models.inception_v3: Add information about expected input size

[sauercrowd]

While using the pretrained inception v3 model I wasnt aware that the input size has to be 299x299, as I figured out after a little bit of try and error and searching.

I hope I didn't miss something in the docs, but if not it might make sense to add a small note in the models.inception_v3 doc that the expected tensor size is different for this model.

[fmassa]

This is currently "documented" in https://github.com/pytorch/vision/blob/master/torchvision/models/inception.py#L77, but I agree that it would be better to have it more visible in the official documentations.
Could you send a PR improving the documentation?

[sauercrowd]

Sure, I take that!

[sauercrowd]

Sorry for the late response! had a bit of an issue with the doc building process to verify it looks properly, but a PR is in place now!

[ekagra-ranjan]

@fmassa I think this can be closed :)

[sauercrowd]

I'll close it 👍