Qwen3.5-35B-A3B-AutoRound-W4A16-RTN

Model Details

This model is a int4 weight-only quantization with group_size 128 and symmetric quantization of Qwen/Qwen3.5-35B-A3B generated by AutoRound. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	Qwen/Qwen3.5-35B-A3B
Quantization Tool	AutoRound
Quantization Scheme	W4A16
Quantized Size	19936 MB

Evaluation Results

Task	Accuracy
hellaswag	0.6202
mmlu	0.8205
mmlu_abstract_algebra	0.6800
mmlu_anatomy	0.8444
mmlu_astronomy	0.9408
mmlu_business_ethics	0.8700
mmlu_clinical_knowledge	0.9132
mmlu_college_biology	0.9375
mmlu_college_chemistry	0.6200
mmlu_college_computer_science	0.7600
mmlu_college_mathematics	0.7100
mmlu_college_medicine	0.8728
mmlu_college_physics	0.7255
mmlu_computer_security	0.8200
mmlu_conceptual_physics	0.9362
mmlu_econometrics	0.7807
mmlu_electrical_engineering	0.8207
mmlu_elementary_mathematics	0.8069
mmlu_formal_logic	0.6825
mmlu_global_facts	0.4900
mmlu_high_school_biology	0.9613
mmlu_high_school_chemistry	0.7931
mmlu_high_school_computer_science	0.8600
mmlu_high_school_european_history	0.8364
mmlu_high_school_geography	0.9444
mmlu_high_school_government_and_politics	0.9845
mmlu_high_school_macroeconomics	0.9026
mmlu_high_school_mathematics	0.5519
mmlu_high_school_microeconomics	0.9664
mmlu_high_school_physics	0.7815
mmlu_high_school_psychology	0.9560
mmlu_high_school_statistics	0.8102
mmlu_high_school_us_history	0.9118
mmlu_high_school_world_history	0.9156
mmlu_human_aging	0.8117
mmlu_human_sexuality	0.9084
mmlu_humanities	0.7433
mmlu_international_law	0.9008
mmlu_jurisprudence	0.8981
mmlu_logical_fallacies	0.8896
mmlu_machine_learning	0.7768
mmlu_management	0.9320
mmlu_marketing	0.9444
mmlu_medical_genetics	0.9500
mmlu_miscellaneous	0.9387
mmlu_moral_disputes	0.8497
mmlu_moral_scenarios	0.5732
mmlu_nutrition	0.8922
mmlu_other	0.8655
mmlu_philosophy	0.8489
mmlu_prehistory	0.9043
mmlu_professional_accounting	0.7589
mmlu_professional_law	0.6525
mmlu_professional_medicine	0.9265
mmlu_professional_psychology	0.8840
mmlu_public_relations	0.7545
mmlu_security_studies	0.8204
mmlu_social_sciences	0.9074
mmlu_sociology	0.9254
mmlu_stem	0.8065
mmlu_us_foreign_policy	0.9300
mmlu_virology	0.5602
mmlu_world_religions	0.9006
piqa	0.8194

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 = "Qwen3.5-35B-A3B-AutoRound-W4A16-RTN"

# 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 Qwen3.5-35B-A3B-AutoRound-W4A16-RTN \
    --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.