biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Bugfix] Fix block size validation (#10938)

Browse Source
This commit is contained in:
chenqianfzh
2024-12-15 16:38:40 -08:00
committed by GitHub
parent da6f409246
commit 69ba344de8
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/engine/arg_utils.py
View File

@@ -425,7 +425,7 @@ class EngineArgs:
parser.add_argument('--block-size', parser.add_argument('--block-size',
type=int, type=int,
default=EngineArgs.block_size, default=EngineArgs.block_size,
choices=[8, 16, 32, 64, 128], choices=[8, 16, 32],
help='Token block size for contiguous chunks of ' help='Token block size for contiguous chunks of '
'tokens. This is ignored on neuron devices and ' 'tokens. This is ignored on neuron devices and '
'set to max-model-len') 'set to max-model-len')