Intel/Qwen3-30B-A3B-Instruct-2507-int4-asym-AutoRound

Model Details

This model is a mixed int4 model with group_size 128 and symmetric quantization of Qwen/Qwen3-30B-A3B-Instruct-2507 generated by intel/auto-round. mlp.gate layers fallback to 16 bits to ensure runing successfully on vLLM.

Please follow the license of the original model.

How To Use

vLLM usage

vllm serve Intel/Qwen3-30B-A3B-Instruct-2507-int4-asym-AutoRound  --max-model-len 16384

INT4 Inference on CPU/Intel GPU/CUDA

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "Intel/Qwen3-30B-A3B-Instruct-2507-int4-asym-AutoRound"

# 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 = "Give me a short introduction to large language model."
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=16384
)
output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist() 

content = tokenizer.decode(output_ids, skip_special_tokens=True)

print("content:", content)

"""
content: A **large language model (LLM)** is a type of artificial intelligence system trained on vast amounts of text data to understand and generate human-like language.
 These models use deep learning techniques, particularly transformer architectures, to analyze patterns in language and produce coherent, contextually relevant responses.
 LLMs can perform a wide range of tasks, such as answering questions, writing essays, translating languages, summarizing text, and even coding. Examples include models like GPT (Generative Pre-trained Transformer) series by OpenAI and Llama by Meta.
 While powerful, LLMs are not perfect—they can sometimes generate inaccurate or biased information and require careful use to ensure ethical and reliable outcomes.

"""

Generate the model

Here is the sample command to reproduce the model

 auto-round --model Qwen/Qwen3-30B-A3B-Instruct-2507/ --enable_torch_compile --nsamples 512 --fp_layers "mlp.gate" --asym --output_dir ./Qwen3-30B-A3B-Instruct-2507-int4

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 link

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