Code-Writer-V2-Obliterated-BF16-AutoRound-W4A16-RTN

Model Details

This model is a int4 weight-only quantization with group_size 128 and symmetric quantization of groxaxo/Code-Writer-V2-Obliterated-BF16 generated by AutoRound. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	groxaxo/Code-Writer-V2-Obliterated-BF16
Quantization Tool	AutoRound
Quantization Scheme	W4A16
Quantized Size	17832 MB

Evaluation Results

Task	Accuracy
hellaswag	0.6695
mmlu	0.8532
mmlu_abstract_algebra	0.7800
mmlu_anatomy	0.8370
mmlu_astronomy	0.9539
mmlu_business_ethics	0.8300
mmlu_clinical_knowledge	0.9321
mmlu_college_biology	0.9861
mmlu_college_chemistry	0.6800
mmlu_college_computer_science	0.8000
mmlu_college_mathematics	0.7000
mmlu_college_medicine	0.8728
mmlu_college_physics	0.8137
mmlu_computer_security	0.8900
mmlu_conceptual_physics	0.9489
mmlu_econometrics	0.7719
mmlu_electrical_engineering	0.8483
mmlu_elementary_mathematics	0.9127
mmlu_formal_logic	0.7619
mmlu_global_facts	0.5900
mmlu_high_school_biology	0.9548
mmlu_high_school_chemistry	0.8670
mmlu_high_school_computer_science	0.9200
mmlu_high_school_european_history	0.8970
mmlu_high_school_geography	0.9343
mmlu_high_school_government_and_politics	0.9896
mmlu_high_school_macroeconomics	0.9205
mmlu_high_school_mathematics	0.6296
mmlu_high_school_microeconomics	0.9664
mmlu_high_school_physics	0.8676
mmlu_high_school_psychology	0.9560
mmlu_high_school_statistics	0.8472
mmlu_high_school_us_history	0.9461
mmlu_high_school_world_history	0.9409
mmlu_human_aging	0.8430
mmlu_human_sexuality	0.9008
mmlu_humanities	0.7943
mmlu_international_law	0.9256
mmlu_jurisprudence	0.8981
mmlu_logical_fallacies	0.8834
mmlu_machine_learning	0.7946
mmlu_management	0.9320
mmlu_marketing	0.9615
mmlu_medical_genetics	0.9600
mmlu_miscellaneous	0.9349
mmlu_moral_disputes	0.8555
mmlu_moral_scenarios	0.7028
mmlu_nutrition	0.9085
mmlu_other	0.8793
mmlu_philosophy	0.8746
mmlu_prehistory	0.9043
mmlu_professional_accounting	0.7908
mmlu_professional_law	0.7027
mmlu_professional_medicine	0.9412
mmlu_professional_psychology	0.8938
mmlu_public_relations	0.8091
mmlu_security_studies	0.8449
mmlu_social_sciences	0.9149
mmlu_sociology	0.9204
mmlu_stem	0.8551
mmlu_us_foreign_policy	0.9500
mmlu_virology	0.5904
mmlu_world_religions	0.9123
piqa	0.8074

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 = "Code-Writer-V2-Obliterated-BF16-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 Code-Writer-V2-Obliterated-BF16-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:

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.