diff --git a/vllm/engine/arg_utils.py b/vllm/engine/arg_utils.py index 49dc3d0a40774..044ded28d15d5 100644 --- a/vllm/engine/arg_utils.py +++ b/vllm/engine/arg_utils.py @@ -257,7 +257,7 @@ def add_cli_args( parser.add_argument('--block-size', type=int, default=EngineArgs.block_size, - choices=[8, 16, 32, 64, 128], + choices=[8, 16, 32, 64, 128, 256], help='Token block size for contiguous chunks of ' 'tokens.')