diff --git a/vllm/envs.py b/vllm/envs.py index 5b4a2010..85d60f32 100644 --- a/vllm/envs.py +++ b/vllm/envs.py @@ -206,6 +206,7 @@ environment_variables: Dict[str, Callable[[], Any]] = { # - "FLASH_ATTN": use FlashAttention # - "XFORMERS": use XFormers # - "ROCM_FLASH": use ROCmFlashAttention + # - "FLASHINFER": use flashinfer "VLLM_ATTENTION_BACKEND": lambda: os.getenv("VLLM_ATTENTION_BACKEND", None),