xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-16 12:25:41 +08:00
Code Issues Packages Projects Releases Wiki Activity

[MISC][V1] Handle exception of current_platform.get_device_name() in arg_utils (#14379)

Browse Source 
Signed-off-by: Cody Yu <hao.yu.cody@gmail.com>
This commit is contained in:
Cody Yu 2025-03-10 17:42:11 -07:00 committed by GitHub
parent 90e88ab756
commit 485afdd3cb
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 13 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
vllm/engine/arg_utils.py
View File

@ -1440,8 +1440,19 @@ class EngineArgs:
# When no user override, set the default values based on the usage
# context.
# Use different default values for different hardware.
from vllm.platforms import current_platform
device_name = current_platform.get_device_name().lower()
# Try to query the device name on the current platform. If it fails,
# it may be because the platform that imports vLLM is not the same
# as the platform that vLLM is running on (e.g. the case of scaling
# vLLM with Ray) and has no GPUs. In this case we use the default
# values for non-H100/H200 GPUs.
try:
from vllm.platforms import current_platform
device_name = current_platform.get_device_name().lower()
except Exception:
# This is only used to set default_max_num_batched_tokens
device_name = "no-device"
if "h100" in device_name or "h200" in device_name:
# For H100 and H200, we use larger default values.
default_max_num_batched_tokens = {