A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 4, 22:30 UTC
[Misc] Reduce attention logging levels (#40086)
Signed-off-by: chaunceyjiang <chaunceyjiang@gmail.com>
Chauncey committed
Apr 21, 2026 at 10:09 UTC
18563f20727ac1e12802fa9fbc6eb295ac0a2ca8
1 file changed
+1
-1
vllm/model_executor/layers/attention/attention.py
+1
-1
@@ -259,7 +259,7 @@ class Attention(nn.Module, AttentionLayerBase):
259
if skip:
260
kv_cache_dtype = "auto"
261
calculate_kv_scales = False
262
- logger.info(
262
+ logger.debug(
263
"Layer %s: kv_cache_dtype=%s, sliding_window=%s",
264
prefix,
265
kv_cache_dtype,