This model is a int4 weight-only quantization with group_size 128 and symmetric quantization of lordx64/Qwable-v2 generated by AutoRound. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	lordx64/Qwable-v2
Quantization Tool	AutoRound
Quantization Scheme	W4A16
Quantized Size	19936 MB

Evaluation Results

Task	Accuracy
hellaswag	0.6319
mmlu	0.8194
mmlu_abstract_algebra	0.5900
mmlu_anatomy	0.8370
mmlu_astronomy	0.9276
mmlu_business_ethics	0.8500
mmlu_clinical_knowledge	0.8906
mmlu_college_biology	0.9375
mmlu_college_chemistry	0.6400
mmlu_college_computer_science	0.7700
mmlu_college_mathematics	0.6000
mmlu_college_medicine	0.8324
mmlu_college_physics	0.6765
mmlu_computer_security	0.8600
mmlu_conceptual_physics	0.9277
mmlu_econometrics	0.7719
mmlu_electrical_engineering	0.8000
mmlu_elementary_mathematics	0.7989
mmlu_formal_logic	0.7143
mmlu_global_facts	0.5400
mmlu_high_school_biology	0.9548
mmlu_high_school_chemistry	0.8128
mmlu_high_school_computer_science	0.8900
mmlu_high_school_european_history	0.8303
mmlu_high_school_geography	0.9394
mmlu_high_school_government_and_politics	0.9793
mmlu_high_school_macroeconomics	0.8872
mmlu_high_school_mathematics	0.5741
mmlu_high_school_microeconomics	0.9664
mmlu_high_school_physics	0.8079
mmlu_high_school_psychology	0.9596
mmlu_high_school_statistics	0.7870
mmlu_high_school_us_history	0.9118
mmlu_high_school_world_history	0.9198
mmlu_human_aging	0.8296
mmlu_human_sexuality	0.8779
mmlu_humanities	0.7539
mmlu_international_law	0.9008
mmlu_jurisprudence	0.8611
mmlu_logical_fallacies	0.8834
mmlu_machine_learning	0.7143
mmlu_management	0.9320
mmlu_marketing	0.9487
mmlu_medical_genetics	0.9200
mmlu_miscellaneous	0.9361
mmlu_moral_disputes	0.8410
mmlu_moral_scenarios	0.5944
mmlu_nutrition	0.8856
mmlu_other	0.8574
mmlu_philosophy	0.8778
mmlu_prehistory	0.9012
mmlu_professional_accounting	0.7092
mmlu_professional_law	0.6688
mmlu_professional_medicine	0.9375
mmlu_professional_psychology	0.8709
mmlu_public_relations	0.7545
mmlu_security_studies	0.8367
mmlu_social_sciences	0.9028
mmlu_sociology	0.9254
mmlu_stem	0.7983
mmlu_us_foreign_policy	0.9400
mmlu_virology	0.5482
mmlu_world_religions	0.9123
piqa	0.8215

How to Use

HF Usage

Step 1: Install AutoRound

pip install auto-round

Step 2: Load and run the quantized model

python

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "Qwable-v2-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

bash

vllm serve Qwable-v2-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.

LeaderboardModel1/Qwable-v2-AutoRound-W4A16-RTN

Qwable-v2-AutoRound-W4A16-RTN

Model Details