HARC-Qwen2.5-7B-Instruct-AutoRound-W4A16-RTN

Model Details

This model is a int4 weight-only quantization with group_size 128 and symmetric quantization of microsoft/HARC-Qwen2.5-7B-Instruct generated by AutoRound. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	microsoft/HARC-Qwen2.5-7B-Instruct
Quantization Tool	AutoRound
Quantization Scheme	W4A16
Quantized Size	5313 MB

Evaluation Results

Task	Accuracy
hellaswag	0.6042
mmlu	0.7089
mmlu_abstract_algebra	0.5400
mmlu_anatomy	0.7407
mmlu_astronomy	0.8224
mmlu_business_ethics	0.7800
mmlu_clinical_knowledge	0.7774
mmlu_college_biology	0.8750
mmlu_college_chemistry	0.5400
mmlu_college_computer_science	0.6400
mmlu_college_mathematics	0.4700
mmlu_college_medicine	0.6879
mmlu_college_physics	0.4706
mmlu_computer_security	0.8200
mmlu_conceptual_physics	0.7277
mmlu_econometrics	0.6316
mmlu_electrical_engineering	0.7241
mmlu_elementary_mathematics	0.6349
mmlu_formal_logic	0.5635
mmlu_global_facts	0.4000
mmlu_high_school_biology	0.8387
mmlu_high_school_chemistry	0.6502
mmlu_high_school_computer_science	0.8300
mmlu_high_school_european_history	0.8364
mmlu_high_school_geography	0.8939
mmlu_high_school_government_and_politics	0.9326
mmlu_high_school_macroeconomics	0.7538
mmlu_high_school_mathematics	0.5333
mmlu_high_school_microeconomics	0.8403
mmlu_high_school_physics	0.5762
mmlu_high_school_psychology	0.8899
mmlu_high_school_statistics	0.6667
mmlu_high_school_us_history	0.8873
mmlu_high_school_world_history	0.8523
mmlu_human_aging	0.7444
mmlu_human_sexuality	0.8168
mmlu_humanities	0.6304
mmlu_international_law	0.8099
mmlu_jurisprudence	0.7685
mmlu_logical_fallacies	0.8466
mmlu_machine_learning	0.5000
mmlu_management	0.8641
mmlu_marketing	0.9274
mmlu_medical_genetics	0.8400
mmlu_miscellaneous	0.8557
mmlu_moral_disputes	0.7630
mmlu_moral_scenarios	0.4525
mmlu_nutrition	0.7941
mmlu_other	0.7596
mmlu_philosophy	0.7396
mmlu_prehistory	0.8025
mmlu_professional_accounting	0.5355
mmlu_professional_law	0.4896
mmlu_professional_medicine	0.7537
mmlu_professional_psychology	0.7467
mmlu_public_relations	0.7455
mmlu_security_studies	0.7633
mmlu_social_sciences	0.8144
mmlu_sociology	0.8806
mmlu_stem	0.6730
mmlu_us_foreign_policy	0.8800
mmlu_virology	0.5542
mmlu_world_religions	0.8480
piqa	0.7938

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 = "HARC-Qwen2.5-7B-Instruct-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 HARC-Qwen2.5-7B-Instruct-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.