Awni Hannun

Thanks for the wonderful article! A few minor typos and comments below: - Edge pooling figure has wrong number of embeddings? - Usually "generalized linear model" refers to the link...

## Proposed changes Enables multi output primitives. Also added `divmod` as a multi-output primitive. Not so much because we need it, but I wanted to test the multi-output functionality. ####...

We check `contiguous` in our saving but that doesn't mean row or col contiguous. This is an issue with safetensors as well. I think it would be nice to support...

Add learning rate schedules

4

Add a few learning rate schedules to `mlx.optimizers`.

Common Installation Problems

11

The docs have more information on [troubleshooting installation](https://ml-explore.github.io/mlx/build/html/install.html). Two of the more common issues are described below (as well as in the docs). # Installing from PyPI A common problem...

Contribute Hugging Face models to the MLX Community

27

We encourage you to join the [MLX Community](https://huggingface.co/mlx-community) on Hugging Face 🤗 and upload new MLX converted models and versions of existing models.

## Proposed changes In general this seemed to be fine (as 0 was output for 0/0) but since it's compiler specific / undefined, add a check for 0 in the...

Move lora example to use the same model format / conversion as `hf_llm`

7

So now you can just run it with an HF repo or convert directly from an HF repo (and optionally upload). Should work with a lot more models out of...

Quantize linear layer in LoRA

1

Forgot to remove this when we added transposed quantize kernels.