类库 › omlx
jundot

jundot/omlx

oMLX是为苹果芯片优化的LLM推理服务器,支持连续批处理和分层KV缓存技术,可通过macOS菜单栏直接管理,实现高效的大语言模型本地部署与推理。

13,613 1,158 13,613 332
在 GitHub 上查看
jundot/omlx

截图

https://cdn.buymeacoffee.com/buttons/v2/default-yellow.png
docs/images/omlx_dashboard.png
docs/images/Screenshot 2026-02-10 at 00.36.32.png
docs/images/Screenshot 2026-02-10 at 00.34.30.png
docs/images/Screenshot 2026-02-10 at 00.45.34.png
docs/images/omlx_hot_cold_cache.png
docs/images/omlx_ChatTemplateKwargs.png
docs/images/ScreenShot_2026-03-14_104350_610.png
docs/images/downloader_omlx.png
docs/images/omlx_integrations.png
docs/images/benchmark_omlx.png
docs/images/Screenshot 2026-02-10 at 00.51.54.png

评论

首页 - Wiki
Copyright © 2011-2026 iteam. Current version is 2.155.2. UTC+08:00, 2026-05-13 15:20
浙ICP备14020137号-1 $访客地图$