Escarda-86M-Base-AutoRound-W4A16-RTN

Model Details

This model is a int4 weight-only quantization with group_size 128 and symmetric quantization of Quazim0t0/Escarda-86M-Base generated by AutoRound. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	Quazim0t0/Escarda-86M-Base
Quantization Tool	AutoRound
Quantization Scheme	W4A16
Quantized Size	142 MB

Evaluation Results

Task	Accuracy
hellaswag	0.2802
mmlu	0.2333
mmlu_abstract_algebra	0.2200
mmlu_anatomy	0.2000
mmlu_astronomy	0.1908
mmlu_business_ethics	0.2700
mmlu_clinical_knowledge	0.2302
mmlu_college_biology	0.2847
mmlu_college_chemistry	0.1700
mmlu_college_computer_science	0.2700
mmlu_college_mathematics	0.2100
mmlu_college_medicine	0.2139
mmlu_college_physics	0.2157
mmlu_computer_security	0.2800
mmlu_conceptual_physics	0.3064
mmlu_econometrics	0.2544
mmlu_electrical_engineering	0.2207
mmlu_elementary_mathematics	0.2328
mmlu_formal_logic	0.2937
mmlu_global_facts	0.1600
mmlu_high_school_biology	0.2000
mmlu_high_school_chemistry	0.1773
mmlu_high_school_computer_science	0.2500
mmlu_high_school_european_history	0.2848
mmlu_high_school_geography	0.2374
mmlu_high_school_government_and_politics	0.2280
mmlu_high_school_macroeconomics	0.2231
mmlu_high_school_mathematics	0.2259
mmlu_high_school_microeconomics	0.2017
mmlu_high_school_physics	0.1788
mmlu_high_school_psychology	0.2349
mmlu_high_school_statistics	0.1574
mmlu_high_school_us_history	0.1912
mmlu_high_school_world_history	0.2616
mmlu_human_aging	0.3004
mmlu_human_sexuality	0.2214
mmlu_humanities	0.2334
mmlu_international_law	0.2562
mmlu_jurisprudence	0.3056
mmlu_logical_fallacies	0.2147
mmlu_machine_learning	0.3125
mmlu_management	0.1650
mmlu_marketing	0.2735
mmlu_medical_genetics	0.2700
mmlu_miscellaneous	0.2605
mmlu_moral_disputes	0.2399
mmlu_moral_scenarios	0.2380
mmlu_nutrition	0.2026
mmlu_other	0.2427
mmlu_philosophy	0.1736
mmlu_prehistory	0.1852
mmlu_professional_accounting	0.2411
mmlu_professional_law	0.2405
mmlu_professional_medicine	0.2243
mmlu_professional_psychology	0.2598
mmlu_public_relations	0.2636
mmlu_security_studies	0.1796
mmlu_social_sciences	0.2333
mmlu_sociology	0.2438
mmlu_stem	0.2239
mmlu_us_foreign_policy	0.2500
mmlu_virology	0.2590
mmlu_world_religions	0.2047
piqa	0.5800

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 = "Escarda-86M-Base-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 Escarda-86M-Base-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.