Solved 🎉

Starting from v1.10.0, torch.nn.CrossEntropy()has an arg label_smoothing=0.0 - API link.

Hi, guys. The type torch.LongTensor of target will hinder the implementation like some methods in reference. So is there a possible to add a Arg: label_smoothing for torch.nn.CrossEntropyLoss(), or maybe simply add the docs to show how to convert the target into one-hot vector to work with torch.nn.CrossEntropyLoss() together, or any other simple ways? Thanks.

cc @ezyang @gchanan @zou3519 @bdhirsh @albanD @mruberry