Nicolas Hug

Results 570 comments of Nicolas Hug

I don't think it has been implemented @joecummings . You can see the difference of what the [current issue template looks like for torchtext](https://github.com/pytorch/text/issues/new?assignees=&labels=&template=bug-report.md) vs [what it would look like...

Thanks for the report @Tomsen1410 . It is a known quirk of bicubic mode on floats (it doesn't happen for bilinear mode). There's no good universal solution here: getting values...

should this just be `--device cpu` ?

Thanks for the PR @claudemircasa . It looks like the 2 existing mirrors are up and running - what was the download issue you were experiencing?

Thanks for the report @rbavery > Are there any current efforts in this direction I should be aware of? No, we haven't been looking at the RPN's support for torch.compile...

I think it's a sphinx thing and basically depends on whether or not the rst file contains a toctree or not. Those that contain a toctree will expand, those that...

See also https://github.com/scikit-learn/scikit-learn/issues/16601 regarding a better example gallery

> I think that the suggestion above would be an intermediate step that would still provide a tangible improvement. Agreed!

Maybe it's just a matter of mentioning in the doc that the left branch is the one that respects the condition and the right branch is the rest?

I think this is a very common convention. It follows naturally from the fact that the conditions are of the form `X[i]