Fara1.5-4B-AutoRound-MXFP8-Tuning

Model Details

This model is a MXFP8 quantization of microsoft/Fara1.5-4B generated by TUNING. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	microsoft/Fara1.5-4B
Quantization Tool	TUNING
Quantization Scheme	MXFP8
Quantized Size	5361 MB

Evaluation Results

Task	Accuracy
hellaswag	0.5511
mmlu	0.7245
mmlu_abstract_algebra	0.5300
mmlu_anatomy	0.7630
mmlu_astronomy	0.8882
mmlu_business_ethics	0.7900
mmlu_clinical_knowledge	0.8000
mmlu_college_biology	0.8819
mmlu_college_chemistry	0.5500
mmlu_college_computer_science	0.7200
mmlu_college_mathematics	0.5600
mmlu_college_medicine	0.7572
mmlu_college_physics	0.5686
mmlu_computer_security	0.7600
mmlu_conceptual_physics	0.8170
mmlu_econometrics	0.6491
mmlu_electrical_engineering	0.8138
mmlu_elementary_mathematics	0.6825
mmlu_formal_logic	0.5873
mmlu_global_facts	0.4000
mmlu_high_school_biology	0.9000
mmlu_high_school_chemistry	0.7291
mmlu_high_school_computer_science	0.8300
mmlu_high_school_european_history	0.8182
mmlu_high_school_geography	0.8737
mmlu_high_school_government_and_politics	0.9119
mmlu_high_school_macroeconomics	0.7667
mmlu_high_school_mathematics	0.4852
mmlu_high_school_microeconomics	0.8655
mmlu_high_school_physics	0.7219
mmlu_high_school_psychology	0.9046
mmlu_high_school_statistics	0.7130
mmlu_high_school_us_history	0.8382
mmlu_high_school_world_history	0.8439
mmlu_human_aging	0.6951
mmlu_human_sexuality	0.8244
mmlu_humanities	0.6400
mmlu_international_law	0.8678
mmlu_jurisprudence	0.7685
mmlu_logical_fallacies	0.8160
mmlu_machine_learning	0.5982
mmlu_management	0.8641
mmlu_marketing	0.9145
mmlu_medical_genetics	0.9000
mmlu_miscellaneous	0.8199
mmlu_moral_disputes	0.7197
mmlu_moral_scenarios	0.4939
mmlu_nutrition	0.7876
mmlu_other	0.7618
mmlu_philosophy	0.7203
mmlu_prehistory	0.8025
mmlu_professional_accounting	0.5816
mmlu_professional_law	0.5163
mmlu_professional_medicine	0.8199
mmlu_professional_psychology	0.7647
mmlu_public_relations	0.7273
mmlu_security_studies	0.7551
mmlu_social_sciences	0.8193
mmlu_sociology	0.8408
mmlu_stem	0.7212
mmlu_us_foreign_policy	0.9000
mmlu_virology	0.5241
mmlu_world_religions	0.8363
piqa	0.7590

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 = "Fara1.5-4B-AutoRound-MXFP8-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 Fara1.5-4B-AutoRound-MXFP8-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.