A JavaScript implementation of Llama 3 using node-mlx, code modified from mlx-examples.

Quantized models can only run on Macs with Apple Silicon.

Download weights (more can be found at mlx-community):

npm install -g @frost-beta/huggingface
huggingface download --to weights mlx-community/Meta-Llama-3-8B-Instruct-8bit

Start chating:

npm install -g llama3
llama3-chat ./weights

This project serves as a demo of node-mlx, and code is intended to keep as simple as possible.

For general purpose LLM modules, please visit 🚧.