gpt-oss-20B-mlx-metal-int4

This is a quantized INT4 model based on Apple MLX Framework gpt-oss-20B. You can deploy it on Apple Silicon devices (M1,M2,M3,M4).

Note: This is unoffical version,just for test and dev.

Installation

pip install -U mlx-lm

Conversion

python -m mlx_lm.convert --hf-path openai/gpt-oss-20b  -q

Samples

from mlx_lm import load, generate

model, tokenizer = load("Your gpt-oss-20B-mlx-metal-int4 Path")

prompt = tokenizer.apply_chat_template(
    [{"role": "user", "content": "can you introduce yourself"}],
    tokenize=False,
    add_generation_prompt=True,
)

response = generate(model, tokenizer, prompt=prompt,max_tokens=1024, verbose=True)