This model is a MXFP4 mixed model of Qwen3.5-397B-A17B tuned by intel/auto-round with llm_compressor format. Please follow the license of the original model.
| Configuration | GSM8K | MMLU | PIQA | HelleSwag | Average | Relative to BF16 |
|---|---|---|---|---|---|---|
| BF16 | 0.9765 | 0.8856 | 0.8303 | 0.7423 | 0.858675 | - |
| MXFP4 | 0.9719 | 0.8809 | 0.8292 | 0.7355 | 0.854375 | 99.5% |
How to Run Locally
vllm serve \
INCModel2/Qwen3.5-397B-A17B-MXFP4-Mixed-CT-AutoRound \
--tensor-parallel-size 4 \
--max-model-len 131072 \
--port 8000
curl -s http://127.0.0.1:8000/v1/chat/completions -H "Content-Type: application/json" -d '{
"model": "INCModel2/Qwen3.5-397B-A17B-MXFP4-Mixed-CT-AutoRound",
"messages": [
{"role":"user","content":"2+3=?"}
],
"max_tokens": 10,
"extra_body": {
"chat_template_kwargs": {
"enable_thinking": true
}
}
}' | python3 -m json.tool
Generate the Model
auto-round Qwen/Qwen3.5-397B-A17B \
--model_free \
--scheme MXFP8 \
--ignore_layers "visual,lm_head,embed_tokens,mlp.gate,mlp.shared_expert_gate,self_attn,linear_attn" \
--layer_config "{mlp.experts:{bits:4,data_type:mx_fp}}" \
--format llm_compressor \
--output_dir "./Qwen3.5-397B-A17B-MXFP4-Mixed-CT-AutoRound"
Ethical Considerations and Limitations
The model can produce factually incorrect output, and should not be relied on to produce factually accurate information. Because of the limitations of the pretrained model and the finetuning datasets, it is possible that this model could generate lewd, biased or otherwise offensive outputs. Therefore, before deploying any applications of the model, developers should perform safety testing.
Caveats and Recommendations
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Here are a couple of useful links to learn more about Intel's AI software:
Disclaimer
The license on this model does not constitute legal advice. We are not responsible for the actions of third parties who use this model. Please consult an attorney before using this model for commercial purposes.
Cite
@article{cheng2023optimize, title={Optimize weight rounding via signed gradient descent for the quantization of llms}, author={Cheng, Wenhua and Zhang, Weiwei and Shen, Haihao and Cai, Yiyang and He, Xin and Lv, Kaokao and Liu, Yi}, journal={arXiv preprint arXiv:2309.05516}, year={2023} } arxiv github