mamba2-130m-4bit

4-bit affine quantization of mlx-community/mamba2-130m, produced with FFAI 0.1.0's ffai convert (mlx-affine format, group_size=64).

Conversion

ffai convert mlx-community/mamba2-130m --bits 4 \
    --upload-repo ekryski/mamba2-130m-4bit

See also

• FFAI — fast Apple Silicon LLM inference. Model.load("ekryski/mamba2-130m-4bit") runs this checkpoint end-to-end.
• FFAI quickstart
• FFAI quantization docs