How to use ariel-pillar/llama-3.2-3b-mlx-verify with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir llama-3.2-3b-mlx-verify ariel-pillar/llama-3.2-3b-mlx-verify
4-bit MLX conversion of Llama-3.2-3B-Instruct for Apple Silicon.
original/ holds the pre-conversion artifacts kept for reproducibility.
original/
This is for a POC demonstration, please don't run.
Quantized
