This model is a int4 weight-only quantization with group_size 128 and symmetric quantization of empero-ai/Qwythos-9B-v2 generated by TUNING. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	empero-ai/Qwythos-9B-v2
Quantization Tool	TUNING
Quantization Scheme	W4A16
Quantized Size	8348 MB

Evaluation Results

Task	Accuracy
hellaswag	0.5744
mmlu	0.7673
mmlu_abstract_algebra	0.6200
mmlu_anatomy	0.7556
mmlu_astronomy	0.9079
mmlu_business_ethics	0.8100
mmlu_clinical_knowledge	0.8264
mmlu_college_biology	0.9306
mmlu_college_chemistry	0.5900
mmlu_college_computer_science	0.7500
mmlu_college_mathematics	0.5400
mmlu_college_medicine	0.7977
mmlu_college_physics	0.6275
mmlu_computer_security	0.8400
mmlu_conceptual_physics	0.8681
mmlu_econometrics	0.6491
mmlu_electrical_engineering	0.8069
mmlu_elementary_mathematics	0.7460
mmlu_formal_logic	0.6349
mmlu_global_facts	0.4400
mmlu_high_school_biology	0.9355
mmlu_high_school_chemistry	0.7783
mmlu_high_school_computer_science	0.8400
mmlu_high_school_european_history	0.8788
mmlu_high_school_geography	0.9444
mmlu_high_school_government_and_politics	0.9585
mmlu_high_school_macroeconomics	0.8462
mmlu_high_school_mathematics	0.5111
mmlu_high_school_microeconomics	0.9076
mmlu_high_school_physics	0.6887
mmlu_high_school_psychology	0.9193
mmlu_high_school_statistics	0.7639
mmlu_high_school_us_history	0.8873
mmlu_high_school_world_history	0.8987
mmlu_human_aging	0.7758
mmlu_human_sexuality	0.8321
mmlu_humanities	0.6861
mmlu_international_law	0.8512
mmlu_jurisprudence	0.8611
mmlu_logical_fallacies	0.8466
mmlu_machine_learning	0.6429
mmlu_management	0.8544
mmlu_marketing	0.9316
mmlu_medical_genetics	0.9100
mmlu_miscellaneous	0.8851
mmlu_moral_disputes	0.7832
mmlu_moral_scenarios	0.4927
mmlu_nutrition	0.8627
mmlu_other	0.8101
mmlu_philosophy	0.8199
mmlu_prehistory	0.8148
mmlu_professional_accounting	0.6312
mmlu_professional_law	0.5834
mmlu_professional_medicine	0.8860
mmlu_professional_psychology	0.8219
mmlu_public_relations	0.6909
mmlu_security_studies	0.7755
mmlu_social_sciences	0.8590
mmlu_sociology	0.9055
mmlu_stem	0.7567
mmlu_us_foreign_policy	0.9000
mmlu_virology	0.5361
mmlu_world_religions	0.8713
piqa	0.7894

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 = "Qwythos-9B-v2-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

bash

vllm serve Qwythos-9B-v2-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.

LeaderboardModel1/Qwythos-9B-v2-AutoRound-W4A16-Tuning

Qwythos-9B-v2-AutoRound-W4A16-Tuning

Model Details