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

[Bugfix] Fix flag name for max_seq_len_to_capture (#4935)

Browse Source 
Signed-off-by: kerthcet <kerthcet@gmail.com>
This commit is contained in:
Kante Yin 2024-05-22 00:30:52 +08:00 committed by GitHub
parent 757b62c495
commit 14772eeb8e
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
vllm/engine/arg_utils.py
View File

@ -341,9 +341,9 @@ class EngineArgs:
help='Maximum context length covered by CUDA ' help='Maximum context length covered by CUDA '
'graphs. When a sequence has context length ' 'graphs. When a sequence has context length '
'larger than this, we fall back to eager mode. ' 'larger than this, we fall back to eager mode. '
'(DEPRECATED. Use --max-seq_len-to-capture instead' '(DEPRECATED. Use --max-seq-len-to-capture instead'
')') ')')
parser.add_argument('--max-seq_len-to-capture', parser.add_argument('--max-seq-len-to-capture',
type=int, type=int,
default=EngineArgs.max_seq_len_to_capture, default=EngineArgs.max_seq_len_to_capture,
help='Maximum sequence length covered by CUDA ' help='Maximum sequence length covered by CUDA '