Matt

Hi @sanjeevk-os - we're getting even more requests for this, so we'd like to try to add it soon! If you're having trouble, just let us know. We can take...

This issue has now been resolved - thank you to @sanjeevk-os for the very clean PR!

I figured this one out! Making a PR.

@hexie1995 This should now be fixed on main! You can install from `main` with `pip install git+https://github.com/huggingface/transformers.git`. It will also be included in the next release, at which point you...

Update: Slightly delayed because there are some breaking architecture changes between the different Falcon checkpoints - I'm merging the various layers and using config variables to switch between the behaviours.

Hey all! The main modeling code should be ready for final review now. Thanks @ArthurZucker for the comprehensive review - it was really helpful! There's one bug left that's causing...

Hi, I'm sorry, I'm not sure how I missed your last comment - this is entirely my fault! Let me investigate the errors you were getting and I'll see if...

Hi @joaocmd, I just took a look now. The cause of the errors is that weight names are not being matched up correctly between the saved checkpoint and the model....

This looks pretty good! Is there a reason to use `Union[torch.Tensor, torch.LongTensor]` instead of just `torch.LongTensor`?

Hi @dewasahu2003, I think in most cases we just annotate those types as `LongTensor`! Your version is probably more correct, but for simplicity just `LongTensor` is fine, since that's what...