This model is a int4 weight-only quantization with group_size 128 and symmetric quantization of ConicCat/Qwen3.5-27B-Writer-V2 generated by TUNING. Please follow the license of the original model.

Quantization Details

Attribute	Value
Base Model	ConicCat/Qwen3.5-27B-Writer-V2
Quantization Tool	TUNING
Quantization Scheme	W4A16
Quantized Size	17832 MB

Evaluation Results

Task	Accuracy
hellaswag	0.6712
mmlu	0.8552
mmlu_abstract_algebra	0.7700
mmlu_anatomy	0.8296
mmlu_astronomy	0.9605
mmlu_business_ethics	0.8500
mmlu_clinical_knowledge	0.9057
mmlu_college_biology	0.9653
mmlu_college_chemistry	0.6900
mmlu_college_computer_science	0.8400
mmlu_college_mathematics	0.7200
mmlu_college_medicine	0.8439
mmlu_college_physics	0.7941
mmlu_computer_security	0.8700
mmlu_conceptual_physics	0.9447
mmlu_econometrics	0.8246
mmlu_electrical_engineering	0.8483
mmlu_elementary_mathematics	0.9127
mmlu_formal_logic	0.7381
mmlu_global_facts	0.6200
mmlu_high_school_biology	0.9484
mmlu_high_school_chemistry	0.8571
mmlu_high_school_computer_science	0.9300
mmlu_high_school_european_history	0.8848
mmlu_high_school_geography	0.9545
mmlu_high_school_government_and_politics	0.9845
mmlu_high_school_macroeconomics	0.9231
mmlu_high_school_mathematics	0.6593
mmlu_high_school_microeconomics	0.9706
mmlu_high_school_physics	0.8344
mmlu_high_school_psychology	0.9523
mmlu_high_school_statistics	0.8657
mmlu_high_school_us_history	0.9510
mmlu_high_school_world_history	0.9409
mmlu_human_aging	0.8475
mmlu_human_sexuality	0.9160
mmlu_humanities	0.8028
mmlu_international_law	0.9339
mmlu_jurisprudence	0.9167
mmlu_logical_fallacies	0.9080
mmlu_machine_learning	0.7946
mmlu_management	0.8932
mmlu_marketing	0.9573
mmlu_medical_genetics	0.9600
mmlu_miscellaneous	0.9387
mmlu_moral_disputes	0.8584
mmlu_moral_scenarios	0.7374
mmlu_nutrition	0.8987
mmlu_other	0.8745
mmlu_philosophy	0.8746
mmlu_prehistory	0.9043
mmlu_professional_accounting	0.7731
mmlu_professional_law	0.7086
mmlu_professional_medicine	0.9485
mmlu_professional_psychology	0.8873
mmlu_public_relations	0.7545
mmlu_security_studies	0.8327
mmlu_social_sciences	0.9155
mmlu_sociology	0.9403
mmlu_stem	0.8557
mmlu_us_foreign_policy	0.9500
mmlu_virology	0.5843
mmlu_world_religions	0.8889
piqa	0.8030

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 = "Qwen3.5-27B-Writer-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 Qwen3.5-27B-Writer-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/Qwen3.5-27B-Writer-V2-AutoRound-W4A16-Tuning

Qwen3.5-27B-Writer-V2-AutoRound-W4A16-Tuning

Model Details