qq1060/sglang
SGLang is a high-performance serving framework for large language models and multimodal models.
SGLang is a high-performance serving framework for large language models and multimodal models.
Train speculative decoding models effortlessly and port them smoothly to SGLang serving.
A unified library for building, evaluating, and storing speculative decoding algorithms for LLM inference in vLLM
support deepstack for ealge3
:star2: Wiki of OI / ICPC for everyone. (某大型游戏线上攻略,内含炫酷算术魔法)
wtfpython的中文翻译/施工结束/ 能力有限,欢迎帮我改进翻译
https://github.com/CyC2018/Interview-Notebook