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 8, 11:15 UTC
[Benchmarks] Reduce ready checker log verbosity (#34349)
Signed-off-by: Tomas Ruiz <tomas.ruiz.te@gmail.com>
Tomas Ruiz committed
Feb 11, 2026 at 21:57 UTC
144d9b7cc8352c5868eb407dd970be94f02b572f
1 file changed
+2
-1
vllm/benchmarks/lib/ready_checker.py
+2
-1
@@ -66,7 +66,8 @@ async def wait_for_endpoint(
66
pbar.close()
67
return output
68
else:
69
- logger.warning("Endpoint is not ready. Error='%s'", output.error)
69
+ err_last_line = str(output.error).rstrip().rsplit("\n", 1)[-1]
70
+ logger.warning("Endpoint is not ready. Error='%s'", err_last_line)
71
except aiohttp.ClientConnectorError:
72
pass
73