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 15, 08:30 UTC
[BugFix] Increase TP execute_model timeout (#18558)
Signed-off-by: Nick Hill <nhill@redhat.com>
Nick Hill committed
May 22, 2025 at 19:22 UTC
93ecb8139c2a13675273835746d025f3ffc61f6d
1 file changed
+1
-1
vllm/v1/executor/multiproc_executor.py
+1
-1
@@ -38,7 +38,7 @@ logger = init_logger(__name__)
38
POLLING_TIMEOUT_MS = 5000
39
POLLING_TIMEOUT_S = POLLING_TIMEOUT_MS // 1000
40
41
-EXECUTE_MODEL_TIMEOUT_S = 40
41
+EXECUTE_MODEL_TIMEOUT_S = 300
42
43
44
class MultiprocExecutor(Executor):