Qwen3-8B-AutoRound-NVFP4-RTN

Model Details

This model is a NVFP4 (NVIDIA FP4) quantization of Qwen/Qwen3-8B generated by AutoRound. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	Qwen/Qwen3-8B
Quantization Tool	AutoRound
Quantization Scheme	NVFP4
Quantized Size	8171 MB

Evaluation Results

Task	Accuracy
hellaswag	0.5584
mmlu	0.7184
mmlu_abstract_algebra	0.5800
mmlu_anatomy	0.7037
mmlu_astronomy	0.8947
mmlu_business_ethics	0.7400
mmlu_clinical_knowledge	0.7774
mmlu_college_biology	0.8542
mmlu_college_chemistry	0.5500
mmlu_college_computer_science	0.7000
mmlu_college_mathematics	0.5600
mmlu_college_medicine	0.7572
mmlu_college_physics	0.6373
mmlu_computer_security	0.8200
mmlu_conceptual_physics	0.8468
mmlu_econometrics	0.6842
mmlu_electrical_engineering	0.7862
mmlu_elementary_mathematics	0.7037
mmlu_formal_logic	0.6270
mmlu_global_facts	0.4300
mmlu_high_school_biology	0.8903
mmlu_high_school_chemistry	0.6798
mmlu_high_school_computer_science	0.8600
mmlu_high_school_european_history	0.8424
mmlu_high_school_geography	0.8586
mmlu_high_school_government_and_politics	0.9223
mmlu_high_school_macroeconomics	0.7718
mmlu_high_school_mathematics	0.4926
mmlu_high_school_microeconomics	0.8950
mmlu_high_school_physics	0.6755
mmlu_high_school_psychology	0.9009
mmlu_high_school_statistics	0.7269
mmlu_high_school_us_history	0.8676
mmlu_high_school_world_history	0.8523
mmlu_human_aging	0.7130
mmlu_human_sexuality	0.8244
mmlu_humanities	0.6225
mmlu_international_law	0.7851
mmlu_jurisprudence	0.7870
mmlu_logical_fallacies	0.8466
mmlu_machine_learning	0.5714
mmlu_management	0.8932
mmlu_marketing	0.9402
mmlu_medical_genetics	0.8200
mmlu_miscellaneous	0.8378
mmlu_moral_disputes	0.7283
mmlu_moral_scenarios	0.3788
mmlu_nutrition	0.7778
mmlu_other	0.7612
mmlu_philosophy	0.7749
mmlu_prehistory	0.8148
mmlu_professional_accounting	0.5957
mmlu_professional_law	0.5026
mmlu_professional_medicine	0.7647
mmlu_professional_psychology	0.7386
mmlu_public_relations	0.7000
mmlu_security_studies	0.7837
mmlu_social_sciences	0.8187
mmlu_sociology	0.8507
mmlu_stem	0.7215
mmlu_us_foreign_policy	0.8800
mmlu_virology	0.5301
mmlu_world_religions	0.8596
piqa	0.7688

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 = "Qwen3-8B-AutoRound-NVFP4-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 Qwen3-8B-AutoRound-NVFP4-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.