Llama-3.1-Nemotron-Nano-VL-8B-V1-4bit

4-bit affine quantization of nvidia/Llama-3.1-Nemotron-Nano-VL-8B-V1, produced with FFAI 0.1.0's ffai convert (mlx-affine format, group_size=64).

Conversion

ffai convert nvidia/Llama-3.1-Nemotron-Nano-VL-8B-V1 --bits 4 \
    --upload-repo ekryski/Llama-3.1-Nemotron-Nano-VL-8B-V1-4bit

See also

• FFAI — fast Apple Silicon LLM inference. Model.load("ekryski/Llama-3.1-Nemotron-Nano-VL-8B-V1-4bit") runs this checkpoint end-to-end.
• FFAI quickstart
• FFAI quantization docs