xlite-dev/Awesome-LLM-Inference

Awesome-LLM-Inference

✅ 近期有维护

📚A curated list of Awesome LLM/VLM Inference Papers with Codes: Flash-Attention, Paged-Attention, WINT8/4, Parallelism, etc.🎉

5,542Stars

443Forks

+4近 7 天 Star

Python · GPL-3.0 · 最近推送 2026-10-20

官方仓库与安装文档 ↗

flash-attention paged-attention tensorrt-llm vllm awesome-llm llm-inference deepseek flash-attention-3 deepseek-v3 minimax-01

Star 历史

2025-09-292026-11-02

查看历史数据

日期	Stars
2026-10-04	5,514
2026-10-05	5,515
2026-10-06	5,516
2026-10-07	5,516
2026-10-08	5,517
2026-10-09	5,517
2026-10-10	5,518
2026-10-11	5,521
2026-10-12	5,522
2026-10-13	5,525
2026-10-14	5,526
2026-10-15	5,527
2026-10-16	5,528
2026-10-17	5,528
2026-10-18	5,528
2026-10-19	5,530
2026-10-20	5,532
2026-10-21	5,533
2026-10-22	5,534
2026-10-23	5,535
2026-10-24	5,536
2026-10-25	5,538
2026-10-26	5,538
2026-10-27	5,538
2026-10-28	5,538
2026-10-29	5,540
2026-10-30	5,540
2026-10-31	5,541
2026-11-01	5,542
2026-11-02	5,542

获取项目

git clone https://github.com/xlite-dev/Awesome-LLM-Inference.git

相关项目与平替候选

同领域不等于功能完全相同，选型前请核对官方文档和许可证。

本地一条命令就能跑起 DeepSeek、Qwen、Gemma 等大模型，自带模型管理和 API，适合离线自托管。

★ 185,625 lobehub

🤯 LobeHub is your Chief Agent Operator, organizing your agents into 7×24 operations by hiring, scheduling, and reporting on your entire AI team.

★ 83,691 unsloth

大语言模型微调提速神器：让 Llama、Mistral 训练提速 2~5 倍并降低 80% 显存消耗。