Awni Hannun

Results 1014 comments of Awni Hannun

> The other option is to make the step a scalar array. I like that option, makes saving easier + no changes to eval. Generally still probably ok if you...

@postmalloc any updates on this?

@angeloskath I finished this PR. Can you take a look? Basically finished @postmalloc suggested implementation with some modifications (mostly around making sure everything fits with compile). The usage is pretty...

> Fixed a type annotation for python 3.8. I wonder 1) if we should be running the tests on the older python version we support (probably yes) 2) if we...

This is a question we get asked a lot. Before I get started on why, let me say we like PyTorch a lot, we use it often, and much of...

> what if instead of making MLX compatible with PyTorch we tried to make cuda compatible with a simplified version of MLX Interesting question. Most of the METAL backend is...

Yes were definitely considering a swift front-end and it's well within scope. Prioritizing it depends on how useful it is to MLX users. If you care to say more about...

We are actively working on a swift front end. I can't give a timeline yet but we like to move quicly :) and it's a top priority for us based...

> I'm eager and happy to personally contribute to any efforts towards a Swift frontend That's awesome and encouraged! > does your statement about active work on a swift front...

>I hope that performance of the c++ code will not be significantly compromised by the introduction of that swift front end. The C++ API will not be affected by the...