From 18563f20727ac1e12802fa9fbc6eb295ac0a2ca8 Mon Sep 17 00:00:00 2001
From: Chauncey <chaunceyjiang@gmail.com>
Date: Tue, 21 Apr 2026 10:09:25 +0800
Subject: [PATCH] [Misc] Reduce attention logging levels (#40086)

Signed-off-by: chaunceyjiang <chaunceyjiang@gmail.com>
---
 vllm/model_executor/layers/attention/attention.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/vllm/model_executor/layers/attention/attention.py b/vllm/model_executor/layers/attention/attention.py
index b2a2295ce46..54f0e1ce5fe 100644
--- a/vllm/model_executor/layers/attention/attention.py
+++ b/vllm/model_executor/layers/attention/attention.py
@@ -259,7 +259,7 @@ class Attention(nn.Module, AttentionLayerBase):
             if skip:
                 kv_cache_dtype = "auto"
                 calculate_kv_scales = False
-            logger.info(
+            logger.debug(
                 "Layer %s: kv_cache_dtype=%s, sliding_window=%s",
                 prefix,
                 kv_cache_dtype,