From 54cf1cae6222ece444673fd083ee7b930d4e1d52 Mon Sep 17 00:00:00 2001 From: Woosuk Kwon Date: Thu, 17 Jul 2025 21:57:02 -0700 Subject: [PATCH] [Misc] Do not print async output warning for v1 (#21151) Signed-off-by: Woosuk Kwon --- vllm/platforms/cuda.py | 2 +- vllm/platforms/rocm.py | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/vllm/platforms/cuda.py b/vllm/platforms/cuda.py index 03f0c15270be3..240724a675a44 100644 --- a/vllm/platforms/cuda.py +++ b/vllm/platforms/cuda.py @@ -99,7 +99,7 @@ class CudaPlatformBase(Platform): @classmethod def is_async_output_supported(cls, enforce_eager: Optional[bool]) -> bool: - if enforce_eager: + if enforce_eager and not envs.VLLM_USE_V1: logger.warning( "To see benefits of async output processing, enable CUDA " "graph. Since, enforce-eager is enabled, async output " diff --git a/vllm/platforms/rocm.py b/vllm/platforms/rocm.py index 04637f5c7aa6f..e9e18d3fe8e4b 100644 --- a/vllm/platforms/rocm.py +++ b/vllm/platforms/rocm.py @@ -299,7 +299,7 @@ class RocmPlatform(Platform): @classmethod def is_async_output_supported(cls, enforce_eager: Optional[bool]) -> bool: - if enforce_eager: + if enforce_eager and not envs.VLLM_USE_V1: logger.warning( "To see benefits of async output processing, enable CUDA " "graph. Since, enforce-eager is enabled, async output "