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

Update Help Text for --gpu-memory-utilization Argument (#2183)

Browse Source
This commit is contained in:
Suhong Moon
2023-12-18 14:33:24 -05:00
committed by GitHub
parent 1b7c791d60
commit 290e015c6c
1 changed files with 7 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
vllm/engine/arg_utils.py
View File

@@ -156,11 +156,13 @@ class EngineArgs:
type=int, type=int,
default=EngineArgs.swap_space, default=EngineArgs.swap_space,
help='CPU swap space size (GiB) per GPU') help='CPU swap space size (GiB) per GPU')
parser.add_argument('--gpu-memory-utilization', parser.add_argument(
type=float, '--gpu-memory-utilization',
default=EngineArgs.gpu_memory_utilization, type=float,
help='the percentage of GPU memory to be used for ' default=EngineArgs.gpu_memory_utilization,
'the model executor') help='the fraction of GPU memory to be used for '
'the model executor, which can range from 0 to 1.'
'If unspecified, will use the default value of 0.9.')
parser.add_argument('--max-num-batched-tokens', parser.add_argument('--max-num-batched-tokens',
type=int, type=int,
default=EngineArgs.max_num_batched_tokens, default=EngineArgs.max_num_batched_tokens,