Agents-A1-AutoRound-W4A16-RTN

Model Details

This model is a int4 weight-only quantization with group_size 128 and symmetric quantization of InternScience/Agents-A1 generated by AutoRound. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	InternScience/Agents-A1
Quantization Tool	AutoRound
Quantization Scheme	W4A16
Quantized Size	19504 MB

Evaluation Results

Task	Accuracy
hellaswag	0.6244
mmlu	0.8183
mmlu_abstract_algebra	0.6900
mmlu_anatomy	0.8519
mmlu_astronomy	0.9408
mmlu_business_ethics	0.8400
mmlu_clinical_knowledge	0.8906
mmlu_college_biology	0.9444
mmlu_college_chemistry	0.6100
mmlu_college_computer_science	0.7200
mmlu_college_mathematics	0.7000
mmlu_college_medicine	0.8555
mmlu_college_physics	0.6569
mmlu_computer_security	0.8000
mmlu_conceptual_physics	0.9191
mmlu_econometrics	0.8246
mmlu_electrical_engineering	0.8414
mmlu_elementary_mathematics	0.8148
mmlu_formal_logic	0.6984
mmlu_global_facts	0.4900
mmlu_high_school_biology	0.9548
mmlu_high_school_chemistry	0.8079
mmlu_high_school_computer_science	0.9100
mmlu_high_school_european_history	0.8727
mmlu_high_school_geography	0.9545
mmlu_high_school_government_and_politics	0.9741
mmlu_high_school_macroeconomics	0.9077
mmlu_high_school_mathematics	0.5630
mmlu_high_school_microeconomics	0.9496
mmlu_high_school_physics	0.7682
mmlu_high_school_psychology	0.9523
mmlu_high_school_statistics	0.8241
mmlu_high_school_us_history	0.9118
mmlu_high_school_world_history	0.9198
mmlu_human_aging	0.8072
mmlu_human_sexuality	0.8931
mmlu_humanities	0.7426
mmlu_international_law	0.9008
mmlu_jurisprudence	0.8704
mmlu_logical_fallacies	0.8957
mmlu_machine_learning	0.7768
mmlu_management	0.9126
mmlu_marketing	0.9359
mmlu_medical_genetics	0.9500
mmlu_miscellaneous	0.9361
mmlu_moral_disputes	0.8324
mmlu_moral_scenarios	0.5687
mmlu_nutrition	0.8824
mmlu_other	0.8574
mmlu_philosophy	0.8489
mmlu_prehistory	0.9012
mmlu_professional_accounting	0.7447
mmlu_professional_law	0.6545
mmlu_professional_medicine	0.9412
mmlu_professional_psychology	0.8873
mmlu_public_relations	0.7455
mmlu_security_studies	0.8245
mmlu_social_sciences	0.9067
mmlu_sociology	0.9154
mmlu_stem	0.8065
mmlu_us_foreign_policy	0.9200
mmlu_virology	0.5422
mmlu_world_religions	0.8889
piqa	0.8156

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 = "Agents-A1-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 Agents-A1-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.