Qwen3-30B-A3B-Gemini-Pro-High-Reasoning-2507-ABLITERATED-UNCENSORED-qx86-hi-mlx

Brainwaves

          arc   arc/e boolq hswag obkqa piqa  wino
qx86-hi   0.404,0.461,0.652,0.678,0.384,0.781,0.637

Qwen3-30B-A3B-Gemini-Pro-High-Reasoning-2507
qx86-hi   0.422,0.474,0.761,0.687,0.382,0.783,0.647

Qwen3-30B-A3B-Thinking-2507-512k
qx6       0.413,0.449,0.677,0.635,0.392,0.767,0.665

This model Qwen3-30B-A3B-Gemini-Pro-High-Reasoning-2507-ABLITERATED-UNCENSORED-qx86-hi-mlx was converted to MLX format from DavidAU/Qwen3-30B-A3B-Gemini-Pro-High-Reasoning-2507-ABLITERATED-UNCENSORED using mlx-lm version 0.30.7.

Use with mlx

pip install mlx-lm

from mlx_lm import load, generate

model, tokenizer = load("Qwen3-30B-A3B-Gemini-Pro-High-Reasoning-2507-ABLITERATED-UNCENSORED-qx86-hi-mlx")

prompt = "hello"

if tokenizer.chat_template is not None:
    messages = [{"role": "user", "content": prompt}]
    prompt = tokenizer.apply_chat_template(
        messages, add_generation_prompt=True, return_dict=False,
    )

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