This model is a int4 weight-only quantization with group_size 128 and symmetric quantization of Qwen/Qwen-AgentWorld-35B-A3B generated by TUNING. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	Qwen/Qwen-AgentWorld-35B-A3B
Quantization Tool	TUNING
Quantization Scheme	W4A16
Quantized Size	19504 MB

Evaluation Results

Task	Accuracy
hellaswag	0.6298
mmlu	0.8280
mmlu_abstract_algebra	0.6900
mmlu_anatomy	0.8444
mmlu_astronomy	0.9211
mmlu_business_ethics	0.8500
mmlu_clinical_knowledge	0.8906
mmlu_college_biology	0.9306
mmlu_college_chemistry	0.6700
mmlu_college_computer_science	0.7800
mmlu_college_mathematics	0.6300
mmlu_college_medicine	0.8381
mmlu_college_physics	0.7255
mmlu_computer_security	0.8500
mmlu_conceptual_physics	0.9319
mmlu_econometrics	0.7895
mmlu_electrical_engineering	0.8414
mmlu_elementary_mathematics	0.8175
mmlu_formal_logic	0.6984
mmlu_global_facts	0.5600
mmlu_high_school_biology	0.9516
mmlu_high_school_chemistry	0.8079
mmlu_high_school_computer_science	0.8900
mmlu_high_school_european_history	0.8424
mmlu_high_school_geography	0.9343
mmlu_high_school_government_and_politics	0.9689
mmlu_high_school_macroeconomics	0.9000
mmlu_high_school_mathematics	0.6111
mmlu_high_school_microeconomics	0.9580
mmlu_high_school_physics	0.8411
mmlu_high_school_psychology	0.9578
mmlu_high_school_statistics	0.8194
mmlu_high_school_us_history	0.9363
mmlu_high_school_world_history	0.9367
mmlu_human_aging	0.8386
mmlu_human_sexuality	0.9008
mmlu_humanities	0.7598
mmlu_international_law	0.9091
mmlu_jurisprudence	0.9259
mmlu_logical_fallacies	0.8896
mmlu_machine_learning	0.7232
mmlu_management	0.9029
mmlu_marketing	0.9316
mmlu_medical_genetics	0.9100
mmlu_miscellaneous	0.9425
mmlu_moral_disputes	0.8555
mmlu_moral_scenarios	0.6089
mmlu_nutrition	0.8922
mmlu_other	0.8648
mmlu_philosophy	0.8682
mmlu_prehistory	0.8981
mmlu_professional_accounting	0.7553
mmlu_professional_law	0.6669
mmlu_professional_medicine	0.9338
mmlu_professional_psychology	0.8840
mmlu_public_relations	0.7636
mmlu_security_studies	0.8367
mmlu_social_sciences	0.9077
mmlu_sociology	0.9303
mmlu_stem	0.8157
mmlu_us_foreign_policy	0.9500
mmlu_virology	0.5904
mmlu_world_religions	0.9064
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 = "Qwen-AgentWorld-35B-A3B-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 Qwen-AgentWorld-35B-A3B-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/Qwen-AgentWorld-35B-A3B-AutoRound-W4A16-Tuning

Qwen-AgentWorld-35B-A3B-AutoRound-W4A16-Tuning

Model Details