gpt-oss-20b-AutoRound-W4A16-RTN

Model Details

This model is a int4 weight-only quantization with group_size 128 and symmetric quantization of openai/gpt-oss-20b generated by AutoRound. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	openai/gpt-oss-20b
Quantization Tool	AutoRound
Quantization Scheme	W4A16
Quantized Size	12020 MB

Evaluation Results

Task	Accuracy
hellaswag	0.2699
mmlu	0.2455
mmlu_abstract_algebra	0.2400
mmlu_anatomy	0.2000
mmlu_astronomy	0.2171
mmlu_business_ethics	0.2800
mmlu_clinical_knowledge	0.2189
mmlu_college_biology	0.2778
mmlu_college_chemistry	0.2600
mmlu_college_computer_science	0.3000
mmlu_college_mathematics	0.2500
mmlu_college_medicine	0.2312
mmlu_college_physics	0.2157
mmlu_computer_security	0.2100
mmlu_conceptual_physics	0.2553
mmlu_econometrics	0.2456
mmlu_electrical_engineering	0.2552
mmlu_elementary_mathematics	0.2328
mmlu_formal_logic	0.2222
mmlu_global_facts	0.2000
mmlu_high_school_biology	0.2161
mmlu_high_school_chemistry	0.2463
mmlu_high_school_computer_science	0.2200
mmlu_high_school_european_history	0.2182
mmlu_high_school_geography	0.2727
mmlu_high_school_government_and_politics	0.2746
mmlu_high_school_macroeconomics	0.3026
mmlu_high_school_mathematics	0.2296
mmlu_high_school_microeconomics	0.3235
mmlu_high_school_physics	0.2384
mmlu_high_school_psychology	0.2404
mmlu_high_school_statistics	0.2037
mmlu_high_school_us_history	0.2745
mmlu_high_school_world_history	0.2321
mmlu_human_aging	0.2422
mmlu_human_sexuality	0.2748
mmlu_humanities	0.2417
mmlu_international_law	0.1901
mmlu_jurisprudence	0.1389
mmlu_logical_fallacies	0.2638
mmlu_machine_learning	0.3036
mmlu_management	0.2621
mmlu_marketing	0.2393
mmlu_medical_genetics	0.2300
mmlu_miscellaneous	0.2554
mmlu_moral_disputes	0.2341
mmlu_moral_scenarios	0.2447
mmlu_nutrition	0.2255
mmlu_other	0.2369
mmlu_philosophy	0.2862
mmlu_prehistory	0.2099
mmlu_professional_accounting	0.2305
mmlu_professional_law	0.2477
mmlu_professional_medicine	0.2096
mmlu_professional_psychology	0.2549
mmlu_public_relations	0.2818
mmlu_security_studies	0.2939
mmlu_social_sciences	0.2684
mmlu_sociology	0.2289
mmlu_stem	0.2372
mmlu_us_foreign_policy	0.2400
mmlu_virology	0.2349
mmlu_world_religions	0.2573
piqa	0.5631

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 = "gpt-oss-20b-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 gpt-oss-20b-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:

• 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.