Support of MLX

[FoPPi]

MLX it's like ollama but only for arm mac's and faster than ollama. Here is a links of project.

Site
GitHub

Docs from HuggingFace

Official examples

[fblissjr]

Would highly recommend supporting MLX. Runs natively on Apple Silicon, is fast, and is likely to be the defacto inference engine for Apple M-series chips. It's come a long, long way. There's an entire community of models with great quantization features: mlx-community (MLX Community)

Now even supports native distributed mode, so you can run inference over multiple Mac devices - https://github.com/ml-explore/mlx-examples/blob/main/llms/mlx_lm/examples/pipeline_generate.py

I'd be happy to help work on this if there's interest in something beyond ollama.

[ICeZer0]

Since Goose does not support LM-studio as an LLM provider I built an Ollama proxy to convert your queries. Its working on MLX models.

Check it out, hope it helps!
https://github.com/Embedded-Nature/ollama-proxy/