Explore live open-source projects and AI models.

Nano vLLM

Based on Nano-vLLM, a simple replication of vLLM with self-contained paged attention and flash attention implementation

FlexAttention based, minimal vllm-style inference engine for fast Gemma 2 inference.

a fun and educational take on vLLM

注释的nano_vllm仓库，并且完成了MiniCPM4的适配以及注册新模型的功能

nano-vllm是开源的一个gpu推理项目，基于开源版本弄的一个ascend npu版本推理小demo，旨在帮助初学者了解推理的整体流程，区别于vllm，nano-vllm体量更小，麻雀虽小五脏俱全，更有助于初学者学习。

Qwen3-TTS with nano vLLM-style optimizations for fast text-to-speech generation. Achieved 3x faster

Nano vLLM with vLLM v1's request scheduling strategy and chunked prefill

nano-vllm 面试导向学习指南 | 20节课程 + 72道面试八股文 + STAR面试法 + 哆啦A梦风格漫画

A lightweight `vLLM-Omni`-style diffusion implementation built around `Wan2.2-TI2V-5B-Diffusers` inspired from nano-vllm