Saved in:
Bibliographic Details
Main Authors: Ma, Yuxin, Du, Lun, Wei, Lanning, Chen, Kun, Xu, Qian, Wang, Kangyu, Feng, Guofeng, Lu, Guoshan, Liu, Lin, Qi, Xiaojing, Zhang, Xinyuan, Tao, Zhen, Feng, Haibo, Jiang, Ziyun, Xu, Ying, Huang, Zenan, Zhuang, Yihong, Xu, Haokai, Hu, Jiaqi, Lan, Zhenzhong, Zhao, Junbo, Li, Jianguo, Zheng, Da
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2510.08666
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909864024866816
author Ma, Yuxin
Du, Lun
Wei, Lanning
Chen, Kun
Xu, Qian
Wang, Kangyu
Feng, Guofeng
Lu, Guoshan
Liu, Lin
Qi, Xiaojing
Zhang, Xinyuan
Tao, Zhen
Feng, Haibo
Jiang, Ziyun
Xu, Ying
Huang, Zenan
Zhuang, Yihong
Xu, Haokai
Hu, Jiaqi
Lan, Zhenzhong
Zhao, Junbo
Li, Jianguo
Zheng, Da
author_facet Ma, Yuxin
Du, Lun
Wei, Lanning
Chen, Kun
Xu, Qian
Wang, Kangyu
Feng, Guofeng
Lu, Guoshan
Liu, Lin
Qi, Xiaojing
Zhang, Xinyuan
Tao, Zhen
Feng, Haibo
Jiang, Ziyun
Xu, Ying
Huang, Zenan
Zhuang, Yihong
Xu, Haokai
Hu, Jiaqi
Lan, Zhenzhong
Zhao, Junbo
Li, Jianguo
Zheng, Da
contents Diffusion-based large language models (dLLMs) have emerged as a promising alternative to autoregressive (AR) LLMs, leveraging denoising-based generation to enable inherent parallelism. Even more and more open-sourced dLLM models emerge, yet their widespread adoption remains constrained by the lack of a standardized and efficient inference framework. We present dInfer, an efficient and extensible framework for dLLM inference. dInfer decomposes the inference pipeline into four modular components--model, diffusion iteration manager, decoding strategy, and KV-cache manager--and integrates novel algorithms for each component alongside system-level optimizations. Through this combination of algorithmic innovations and system enhancements, dInfer achieves substantial efficiency gains without compromising output quality on LLaDA-MoE. At batch size 1, it surpasses 1,100 tokens per second on HumanEval and averages over 800 tokens per second across six benchmarks on $8\times$ H800 GPUs. Compared to prior systems, dInfer delivers a $10\times$ speedup over Fast-dLLM while maintaining similar model performance. Even compared to the AR model (with a comparable number of activation parameters and performance) QWen2.5-3B, which is highly optimized with the latest vLLM inference engine, dInfer still delivers a $2$-$3\times$ speedup. The implementation of dInfer is open-sourced at https://github.com/inclusionAI/dInfer.
format Preprint
id arxiv_https___arxiv_org_abs_2510_08666
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle dInfer: An Efficient Inference Framework for Diffusion Language Models
Ma, Yuxin
Du, Lun
Wei, Lanning
Chen, Kun
Xu, Qian
Wang, Kangyu
Feng, Guofeng
Lu, Guoshan
Liu, Lin
Qi, Xiaojing
Zhang, Xinyuan
Tao, Zhen
Feng, Haibo
Jiang, Ziyun
Xu, Ying
Huang, Zenan
Zhuang, Yihong
Xu, Haokai
Hu, Jiaqi
Lan, Zhenzhong
Zhao, Junbo
Li, Jianguo
Zheng, Da
Computation and Language
Artificial Intelligence
Diffusion-based large language models (dLLMs) have emerged as a promising alternative to autoregressive (AR) LLMs, leveraging denoising-based generation to enable inherent parallelism. Even more and more open-sourced dLLM models emerge, yet their widespread adoption remains constrained by the lack of a standardized and efficient inference framework. We present dInfer, an efficient and extensible framework for dLLM inference. dInfer decomposes the inference pipeline into four modular components--model, diffusion iteration manager, decoding strategy, and KV-cache manager--and integrates novel algorithms for each component alongside system-level optimizations. Through this combination of algorithmic innovations and system enhancements, dInfer achieves substantial efficiency gains without compromising output quality on LLaDA-MoE. At batch size 1, it surpasses 1,100 tokens per second on HumanEval and averages over 800 tokens per second across six benchmarks on $8\times$ H800 GPUs. Compared to prior systems, dInfer delivers a $10\times$ speedup over Fast-dLLM while maintaining similar model performance. Even compared to the AR model (with a comparable number of activation parameters and performance) QWen2.5-3B, which is highly optimized with the latest vLLM inference engine, dInfer still delivers a $2$-$3\times$ speedup. The implementation of dInfer is open-sourced at https://github.com/inclusionAI/dInfer.
title dInfer: An Efficient Inference Framework for Diffusion Language Models
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2510.08666