gemma-4-12B-AutoRound-W4A16-RTN

Model Details

This model is a int4 weight-only quantization with group_size 128 and symmetric quantization of google/gemma-4-12B generated by AutoRound. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	google/gemma-4-12B
Quantization Tool	AutoRound
Quantization Scheme	W4A16
Quantized Size	7422 MB

Evaluation Results

Task	Accuracy
hellaswag	0.5921
mmlu	0.6796
mmlu_abstract_algebra	0.3500
mmlu_anatomy	0.6963
mmlu_astronomy	0.7895
mmlu_business_ethics	0.7300
mmlu_clinical_knowledge	0.7736
mmlu_college_biology	0.7917
mmlu_college_chemistry	0.4900
mmlu_college_computer_science	0.5700
mmlu_college_mathematics	0.3200
mmlu_college_medicine	0.6994
mmlu_college_physics	0.4412
mmlu_computer_security	0.7500
mmlu_conceptual_physics	0.6766
mmlu_econometrics	0.5088
mmlu_electrical_engineering	0.6828
mmlu_elementary_mathematics	0.5265
mmlu_formal_logic	0.4444
mmlu_global_facts	0.4100
mmlu_high_school_biology	0.8355
mmlu_high_school_chemistry	0.6158
mmlu_high_school_computer_science	0.7300
mmlu_high_school_european_history	0.7636
mmlu_high_school_geography	0.8788
mmlu_high_school_government_and_politics	0.9067
mmlu_high_school_macroeconomics	0.7103
mmlu_high_school_mathematics	0.4111
mmlu_high_school_microeconomics	0.8277
mmlu_high_school_physics	0.4768
mmlu_high_school_psychology	0.8752
mmlu_high_school_statistics	0.6574
mmlu_high_school_us_history	0.8431
mmlu_high_school_world_history	0.8819
mmlu_human_aging	0.7668
mmlu_human_sexuality	0.8168
mmlu_humanities	0.6051
mmlu_international_law	0.8347
mmlu_jurisprudence	0.7963
mmlu_logical_fallacies	0.8344
mmlu_machine_learning	0.5625
mmlu_management	0.8252
mmlu_marketing	0.9103
mmlu_medical_genetics	0.7700
mmlu_miscellaneous	0.8314
mmlu_moral_disputes	0.7803
mmlu_moral_scenarios	0.2413
mmlu_nutrition	0.7810
mmlu_other	0.7444
mmlu_philosophy	0.7460
mmlu_prehistory	0.7747
mmlu_professional_accounting	0.5390
mmlu_professional_law	0.5482
mmlu_professional_medicine	0.7279
mmlu_professional_psychology	0.7582
mmlu_public_relations	0.6909
mmlu_security_studies	0.7592
mmlu_social_sciences	0.7995
mmlu_sociology	0.8806
mmlu_stem	0.6099
mmlu_us_foreign_policy	0.9200
mmlu_virology	0.5241
mmlu_world_religions	0.8830
piqa	0.8030

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 = "gemma-4-12B-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 gemma-4-12B-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.