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 16, 23:15 UTC
[Frontend][Bugfix] Fix preemption_mode -> preemption-mode for CLI arg in arg_utils.py (#5688)
Michael Goin committed
Jun 19, 2024 at 14:41 UTC
afed90a0344b1b0ce6aae46efc630adb489ec769
1 file changed
+1
-1
vllm/engine/arg_utils.py
+1
-1
@@ -577,7 +577,7 @@ class EngineArgs:
577
'This should be a JSON string that will be '
578
'parsed into a dictionary.')
579
parser.add_argument(
580
- '--preemption_mode',
580
+ '--preemption-mode',
581
type=str,
582
default=None,
583
help='If \'recompute\', the engine performs preemption by block '