Amrit Sahu
> Hi, just seeing the merge messed up the commit history. There are 377 changes, which is impossible for the review and merge the PR into `main`. > > I...
Hi @alaradirik , can you review the changes?
Hi @Narsil, Sure the output labels are taken **exactly** from the input text_queries. The reason of naming it "text_queries" instead of "candidate_labels" as in case of zero-shot-image-classification is that, in...
HI @Narsil @alaradirik, kindly review the changes
@guillaumekln so i was trying to enable batching support. Here are some points and blockers I feel. Would love to hear your thoughts. - The transcribe.py needs to handle multiple...
@guillaumekln thanks for responding. Am aware of the position_ids. This can be created on fly via prefix sums on mask. ``` mask_binary = torch.exp(mask) position_ids = mask_binary.long().cumsum(-1) - 1 position_ids.masked_fill_(mask_binary...
@chenyuxyz is there some documentation on `rangeify`? as per my debugging I think this has something to do with `kernelize -> get_rangeify_map`