license: apache-2.0
language:
- en
base_model: LibraxisAI/Huihui-Qwen3.6-35B-A3B-Claude-4.7-Opus-abliterated-vmlx-mxfp8
library_name: mlx
pipeline_tag: image-text-to-text
tags: - image-text-to-text
- vision
- multimodal
- vmlx
- vlm
- reasoning
- distillation
- chain-of-thought
- qwen
- qwen3.6
- mixture-of-experts
- moe
- lora
- unsloth
- abliterated
- uncensored
- mlx
- apple-silicon
- huihui
- quantized
- mxfp8
- mlx
- mlx-my-repo
inference: false
widget: - text: Summarize the operational risks in this deployment plan.
example_title: Reasoning prompt
porschefreak/Huihui-Qwen3.6-35B-A3B-Claude-4.7-Opus-abliterated-vmlx-mxfp8-mlx-4Bit
The Model porschefreak/Huihui-Qwen3.6-35B-A3B-Claude-4.7-Opus-abliterated-vmlx-mxfp8-mlx-4Bit was converted to MLX format from LibraxisAI/Huihui-Qwen3.6-35B-A3B-Claude-4.7-Opus-abliterated-vmlx-mxfp8 using mlx-lm version 0.31.2.
Use with mlx
pip install mlx-lm
from mlx_lm import load, generate
model, tokenizer = load("porschefreak/Huihui-Qwen3.6-35B-A3B-Claude-4.7-Opus-abliterated-vmlx-mxfp8-mlx-4Bit")
prompt="hello"
if hasattr(tokenizer, "apply_chat_template") and tokenizer.chat_template is not None:
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
response = generate(model, tokenizer, prompt=prompt, verbose=True)