Qwen3-0.6B-AutoRound-W4A16-Tuning

Model Details

This model is a int4 weight-only quantization with group_size 128 and symmetric quantization of Qwen/Qwen3-0.6B generated by TUNING. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	Qwen/Qwen3-0.6B
Quantization Tool	TUNING
Quantization Scheme	W4A16
Quantized Size	515 MB

Evaluation Results

Task	Accuracy
hellaswag	0.3653
mmlu	0.4012
mmlu_abstract_algebra	0.3400
mmlu_anatomy	0.3778
mmlu_astronomy	0.4079
mmlu_business_ethics	0.3900
mmlu_clinical_knowledge	0.4038
mmlu_college_biology	0.4236
mmlu_college_chemistry	0.3800
mmlu_college_computer_science	0.3500
mmlu_college_mathematics	0.2300
mmlu_college_medicine	0.3988
mmlu_college_physics	0.3431
mmlu_computer_security	0.5100
mmlu_conceptual_physics	0.3532
mmlu_econometrics	0.2807
mmlu_electrical_engineering	0.4345
mmlu_elementary_mathematics	0.3624
mmlu_formal_logic	0.3492
mmlu_global_facts	0.2000
mmlu_high_school_biology	0.4839
mmlu_high_school_chemistry	0.3300
mmlu_high_school_computer_science	0.4200
mmlu_high_school_european_history	0.5515
mmlu_high_school_geography	0.4495
mmlu_high_school_government_and_politics	0.5130
mmlu_high_school_macroeconomics	0.3564
mmlu_high_school_mathematics	0.3185
mmlu_high_school_microeconomics	0.3740
mmlu_high_school_physics	0.2053
mmlu_high_school_psychology	0.5982
mmlu_high_school_statistics	0.3935
mmlu_high_school_us_history	0.5000
mmlu_high_school_world_history	0.5443
mmlu_human_aging	0.4260
mmlu_human_sexuality	0.4275
mmlu_humanities	0.3656
mmlu_international_law	0.5207
mmlu_jurisprudence	0.4259
mmlu_logical_fallacies	0.4785
mmlu_machine_learning	0.2500
mmlu_management	0.5534
mmlu_marketing	0.5684
mmlu_medical_genetics	0.4400
mmlu_miscellaneous	0.4687
mmlu_moral_disputes	0.3584
mmlu_moral_scenarios	0.2447
mmlu_nutrition	0.4673
mmlu_other	0.4265
mmlu_philosophy	0.3826
mmlu_prehistory	0.4105
mmlu_professional_accounting	0.3050
mmlu_professional_law	0.3240
mmlu_professional_medicine	0.3676
mmlu_professional_psychology	0.4085
mmlu_public_relations	0.4727
mmlu_security_studies	0.4980
mmlu_social_sciences	0.4634
mmlu_sociology	0.5871
mmlu_stem	0.3685
mmlu_us_foreign_policy	0.5400
mmlu_virology	0.3916
mmlu_world_religions	0.4386
piqa	0.6649

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 = "Qwen3-0.6B-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 Qwen3-0.6B-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:

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.