sgl-project / sglang

SGLang is a structured generation language designed for large language models (LLMs). It makes your interaction with models faster and more controllable.
Apache License 2.0
2.75k stars 176 forks source link

Update flashinfer to 0.0.5 #554

Closed merrymercy closed 1 week ago

merrymercy commented 1 week ago

flashinfer 0.0.5 will come with new feature, new behavior and some API changes. This PR fixes the compatibility.