Tess-4-9B-AutoRound-W4A16-Tuning

Model Details

This model is a int4 weight-only quantization with group_size 128 and symmetric quantization of migtissera/Tess-4-9B generated by TUNING. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	migtissera/Tess-4-9B
Quantization Tool	TUNING
Quantization Scheme	W4A16
Quantized Size	8348 MB

Evaluation Results

Task	Accuracy
hellaswag	0.5866
mmlu	0.7611
mmlu_abstract_algebra	0.5800
mmlu_anatomy	0.7926
mmlu_astronomy	0.9079
mmlu_business_ethics	0.7900
mmlu_clinical_knowledge	0.8377
mmlu_college_biology	0.9306
mmlu_college_chemistry	0.5900
mmlu_college_computer_science	0.7300
mmlu_college_mathematics	0.6300
mmlu_college_medicine	0.7919
mmlu_college_physics	0.6275
mmlu_computer_security	0.8300
mmlu_conceptual_physics	0.8553
mmlu_econometrics	0.6930
mmlu_electrical_engineering	0.8000
mmlu_elementary_mathematics	0.7778
mmlu_formal_logic	0.6349
mmlu_global_facts	0.4600
mmlu_high_school_biology	0.9323
mmlu_high_school_chemistry	0.7635
mmlu_high_school_computer_science	0.8500
mmlu_high_school_european_history	0.8485
mmlu_high_school_geography	0.9141
mmlu_high_school_government_and_politics	0.9585
mmlu_high_school_macroeconomics	0.8410
mmlu_high_school_mathematics	0.5074
mmlu_high_school_microeconomics	0.9118
mmlu_high_school_physics	0.6887
mmlu_high_school_psychology	0.9303
mmlu_high_school_statistics	0.7546
mmlu_high_school_us_history	0.9020
mmlu_high_school_world_history	0.9198
mmlu_human_aging	0.7937
mmlu_human_sexuality	0.8626
mmlu_humanities	0.6642
mmlu_international_law	0.8760
mmlu_jurisprudence	0.8333
mmlu_logical_fallacies	0.8344
mmlu_machine_learning	0.6518
mmlu_management	0.8350
mmlu_marketing	0.9444
mmlu_medical_genetics	0.9000
mmlu_miscellaneous	0.8876
mmlu_moral_disputes	0.7803
mmlu_moral_scenarios	0.3855
mmlu_nutrition	0.8464
mmlu_other	0.8104
mmlu_philosophy	0.8039
mmlu_prehistory	0.8395
mmlu_professional_accounting	0.6277
mmlu_professional_law	0.5776
mmlu_professional_medicine	0.8676
mmlu_professional_psychology	0.8121
mmlu_public_relations	0.6909
mmlu_security_studies	0.7837
mmlu_social_sciences	0.8606
mmlu_sociology	0.9055
mmlu_stem	0.7599
mmlu_us_foreign_policy	0.9100
mmlu_virology	0.5602
mmlu_world_religions	0.8655
piqa	0.7927

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 = "Tess-4-9B-AutoRound-W4A16-Tuning"

# 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 Tess-4-9B-AutoRound-W4A16-Tuning \
    --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.