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 14, 05:30 UTC
[UX] Change kv-cache-memory log level to debug (#25479)
Signed-off-by: Michael Goin <mgoin64@gmail.com>
Michael Goin committed
Sep 23, 2025 at 11:01 UTC
2c58742dff8613a3bd7496f2008ce927e18d38d1
1 file changed
+1
-1
vllm/v1/worker/gpu_worker.py
+1
-1
@@ -389,7 +389,7 @@ class Worker(WorkerBase):
389
f"utilize gpu memory. Current kv cache memory in use is "
390
f"{int(self.available_kv_cache_memory_bytes)} bytes.")
391
392
- logger.info(msg)
392
+ logger.debug(msg)
393
394
# Warm up sampler and preallocate memory buffer for logits and other
395
# sampling related tensors of max possible shape to avoid memory