Qwen3.6-27B-AutoRound-W4A16-Tuning

Model Details

This model is a int4 weight-only quantization with group_size 128 and symmetric quantization of Qwen/Qwen3.6-27B generated by TUNING. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	Qwen/Qwen3.6-27B
Quantization Tool	TUNING
Quantization Scheme	W4A16
Quantized Size	18117 MB

Evaluation Results

Task	Accuracy
hellaswag	0.6388
mmlu	0.8461
mmlu_abstract_algebra	0.7400
mmlu_anatomy	0.8519
mmlu_astronomy	0.9342
mmlu_business_ethics	0.8400
mmlu_clinical_knowledge	0.8717
mmlu_college_biology	0.9583
mmlu_college_chemistry	0.6600
mmlu_college_computer_science	0.8200
mmlu_college_mathematics	0.7500
mmlu_college_medicine	0.8555
mmlu_college_physics	0.7157
mmlu_computer_security	0.8500
mmlu_conceptual_physics	0.9447
mmlu_econometrics	0.7807
mmlu_electrical_engineering	0.8276
mmlu_elementary_mathematics	0.8915
mmlu_formal_logic	0.8095
mmlu_global_facts	0.6200
mmlu_high_school_biology	0.9484
mmlu_high_school_chemistry	0.8374
mmlu_high_school_computer_science	0.9000
mmlu_high_school_european_history	0.8909
mmlu_high_school_geography	0.9242
mmlu_high_school_government_and_politics	0.9741
mmlu_high_school_macroeconomics	0.9410
mmlu_high_school_mathematics	0.6593
mmlu_high_school_microeconomics	0.9538
mmlu_high_school_physics	0.8278
mmlu_high_school_psychology	0.9523
mmlu_high_school_statistics	0.8796
mmlu_high_school_us_history	0.9167
mmlu_high_school_world_history	0.9409
mmlu_human_aging	0.8341
mmlu_human_sexuality	0.9160
mmlu_humanities	0.7966
mmlu_international_law	0.9256
mmlu_jurisprudence	0.8981
mmlu_logical_fallacies	0.9080
mmlu_machine_learning	0.7589
mmlu_management	0.8641
mmlu_marketing	0.9359
mmlu_medical_genetics	0.9500
mmlu_miscellaneous	0.9221
mmlu_moral_disputes	0.7948
mmlu_moral_scenarios	0.7363
mmlu_nutrition	0.9020
mmlu_other	0.8632
mmlu_philosophy	0.8617
mmlu_prehistory	0.8673
mmlu_professional_accounting	0.7872
mmlu_professional_law	0.7145
mmlu_professional_medicine	0.9412
mmlu_professional_psychology	0.8693
mmlu_public_relations	0.7818
mmlu_security_studies	0.8122
mmlu_social_sciences	0.9067
mmlu_sociology	0.9353
mmlu_stem	0.8440
mmlu_us_foreign_policy	0.9200
mmlu_virology	0.5542
mmlu_world_religions	0.8947
piqa	0.8134

How to Use

HF Usage

Step 1: Install AutoRound

pip install auto-round

Step 2: Load and run the quantized model

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "Qwen3.6-27B-AutoRound-W4A16-Tuning"

# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype="auto", device_map="auto")

# prepare the model input
prompt = "Write a quick sort algorithm."
messages = [{"role": "user", "content": prompt}]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True,
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)

# conduct text completion
generated_ids = model.generate(**model_inputs, max_new_tokens=512)
output_ids = generated_ids[0][len(model_inputs.input_ids[0]) :].tolist()

content = tokenizer.decode(output_ids, skip_special_tokens=True)
print("content:", content)

VLLM Usage

vllm serve Qwen3.6-27B-AutoRound-W4A16-Tuning \
    --trust-remote-code \
    --dtype bfloat16 \
    --tensor_parallel_size 1

If you encounter any issues, feel free to open an issue on the AutoRound GitHub repo or provide feedback on the Low-Bit Open LLM Leaderboard.

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:

• Intel Neural Compressor
• AutoRound

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

---

This model is part of the Intel Low-Bit Open LLM Leaderboard initiative.