Ring-flash-linear-2.0-GPTQ-int4

Quantized Ring-Linear-2.0

Introduction

To enable deployment of Ring-Linear-2.0 on memory-constrained devices, we release quantized weights using the GPTQ INT4 format. Additionally, we evaluate the online FP8 quantization performance of Ring-Linear-2.0 models, which closely approaches that of BF16 precision.

Model Downloads

Quickstart

🚀 vLLM

Environment Preparation

Since the Pull Request (PR) has not been submitted to the vLLM community at this stage, please prepare the environment by following the steps below.

First, create a Conda environment with Python 3.10 and CUDA 12.8:

conda create -n vllm python=3.10
conda activate vllm

Next, install our vLLM wheel package:

pip install https://media.githubusercontent.com/media/zheyishine/vllm_whl/refs/heads/main/vllm-0.8.5.post2.dev28%2Bgd327eed71.cu128-cp310-cp310-linux_x86_64.whl --force-reinstall

Finally, install compatible versions of transformers after vLLM is installed:

pip install transformers==4.51.1 

Offline Inference

from transformers import AutoTokenizer
from vllm import LLM, SamplingParams

if __name__ == '__main__':
    tokenizer = AutoTokenizer.from_pretrained("inclusionAI/Ring-flash-linear-2.0-GPTQ-int4")
    
    sampling_params = SamplingParams(temperature=0.6, top_p=1.0, max_tokens=16384)

    # use `max_num_seqs=1` without concurrency
    llm = LLM(model="inclusionAI/Ring-flash-linear-2.0-GPTQ-int4", dtype='auto', enable_prefix_caching=False, max_num_seqs=128)
    
    
    prompt = "Give me a short introduction to large language models."
    messages = [
        {"role": "user", "content": prompt}
    ]
    
    text = tokenizer.apply_chat_template(
        messages,
        tokenize=False,
        add_generation_prompt=True
    )
    outputs = llm.generate([text], sampling_params)
    for output in outputs:
        print(output.outputs[0].text)

Online Inference

vllm serve inclusionAI/Ring-flash-linear-2.0-GPTQ-int4 \
              --tensor-parallel-size 2 \
              --pipeline-parallel-size 1 \
              --gpu-memory-utilization 0.90 \
              --max-num-seqs 128 \
              --no-enable-prefix-caching
              --api-key your-api-key

Evaluation

We evaluate the INT4 and FP8 quantized models using several datasets. The FP8 quantization is applied via the quantization="fp8" argument in SGLang or vLLM.

Ring-mini-linear-2.0

Ring-flash-linear-2.0

License

This code repository is licensed under the MIT License.

Citation

@misc{lingteam2025attentionmattersefficienthybrid,
      title={Every Attention Matters: An Efficient Hybrid Architecture for Long-Context Reasoning}, 
      author={Ling Team and Bin Han and Caizhi Tang and Chen Liang and Donghao Zhang and Fan Yuan and Feng Zhu and Jie Gao and Jingyu Hu and Longfei Li and Meng Li and Mingyang Zhang and Peijie Jiang and Peng Jiao and Qian Zhao and Qingyuan Yang and Wenbo Shen and Xinxing Yang and Yalin Zhang and Yankun Ren and Yao Zhao and Yibo Cao and Yixuan Sun and Yue Zhang and Yuchen Fang and Zibin Lin and Zixuan Cheng and Jun Zhou},
      year={2025},
      eprint={2510.19338},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2510.19338}, 
}