--- title: Qwen3-Next-80B-A3B-Instruct-int4-AutoRound canonical_url: "https://www.modelscope.ai/models/Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound" md_url: "https://www.modelscope.ai/models/Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound.md" repository: Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound last_updated: 2026-04-17 license: apache-2.0 pipeline_tag: text-generation tasks: - text-generation model_type: - qwen3_next architectures: - Qwen3NextForCausalLM base_model: - Qwen/Qwen3-Next-80B-A3B-Instruct base_model_relation: quantized parameters: 11.5B tensor_type: - I32 - F16 - BF16 library_name: - pytorch - transformer - safetensors frameworks: - pytorch downloads: 139 stars: 0 --- # Qwen3-Next-80B-A3B-Instruct-int4-AutoRound > Qwen3-Next-80B-A3B-Instruct-int4-AutoRound - An open-source model by Intel on ModelScope. This model is a int4 model with groupsize 128 and symmetric quantization of Qwen/Qwen3-Next-80B-A3B-Instruct generated by intel/auto-round. Please follow the license of… Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound is a 11.5B-parameter text-generation model on ModelScope. licensed under apache-2.0. derived from Qwen/Qwen3-Next-80B-A3B-Instruct. - **Repository**: Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound - **License**: apache-2.0 - **Tasks**: text-generation - **Parameters**: 11.5B - **Base model**: Qwen/Qwen3-Next-80B-A3B-Instruct - **Downloads**: 139 - **Stars**: 0 - **Last updated**: 2026-04-17 Source: https://www.modelscope.ai/models/Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound --- ## Model Details This model is a int4 model with group_size 128 and symmetric quantization of [Qwen/Qwen3-Next-80B-A3B-Instruct](https://huggingface.co/Qwen/Qwen3-Next-80B-A3B-Instruct) generated by [intel/auto-round](https://github.com/intel/auto-round). Please follow the license of the original model. ## How To Use For vllm, this pr is required https://github.com/vllm-project/vllm/pull/24818 ### INT4 Inference ```python from transformers import AutoModelForCausalLM, AutoTokenizer import transformers import torch quantized_model_dir = "Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound" # load the tokenizer and the model tokenizer = AutoTokenizer.from_pretrained(model_name) model = AutoModelForCausalLM.from_pretrained( model_name, 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=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) """ 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 learn patterns, grammar, context, and reasoning from billions of words, enabling them to answer questions, write essays, translate languages, code, and even engage in conversation. Popular examples include OpenAI’s GPT series, Google’s Gemini, and Meta’s Llama. LLMs are foundational to many modern AI applications, from chatbots to content creation tools, though they require careful use due to potential biases, inaccuracies, and ethical concerns. """ ``` ### vLLM The following command can be used to create an API endpoint at `http://localhost:8000/v1` with maximum context length 256K tokens. ```shell vllm serve Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound --port 8000 --max-model-len 262144 ``` The following command is recommended for MTP with the rest settings the same as above: ```shell vllm serve Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound --port 8000 --max-model-len 262144 --speculative-config '{"method":"qwen3_next_mtp","num_speculative_tokens":2}' ``` ```bash curl -noproxy '*' http://localhost::8000/v1/chat/completions \ -H "Content-Type: application/json" \ -d '{ "messages": [ {"role": "user", "content": "Give me a short introduction to large language model."} ], "max_tokens": 1024 }' # "content": # "A large language model (LLM) is a type of artificial intelligence system trained on vast amounts of text data to understand, generate, and manipulate human language. These models use deep learning architectures—often based on the transformer network—to predict the next word in a sequence, enabling them to perform tasks like answering questions, writing essays, translating languages, and even coding. LLMs, such as GPT, Gemini, and Claude, learn patterns and relationships in language without explicit programming, allowing them to produce human-like responses across a wide range of topics. While powerful, they don’t “understand” language in the human sense and can sometimes generate plausible-sounding but incorrect or biased information.", ``` ### Generate the model ```bash AR_DISABLE_COPY_MTP_WEIGHTS=1 auto_round --model Qwen/Qwen3-Next-80B-A3B-Instruct --scheme W4A16 --output_dir tmp_autoround ``` ## Evaluate Results | benchmark | n-shot | backend | Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound | Qwen/Qwen3-Next-80B-A3B-Instruct | | :-------: | :----: | :-----: | :----------------------------------------------: | :------------------------------: | | gsm8k | 5 | vllm | 0.8643 | 0.8074 | | mmlu_pro | 5 | vllm | 0.7570 | 0.7621 | ## 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](https://github.com/intel/neural-compressor) ## 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](https://arxiv.org/abs/2309.05516) [github](https://github.com/intel/auto-round)