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

[Bugfix] Enable V1 usage stats (#16986)

Browse Source 
Signed-off-by: mgoin <mgoin64@gmail.com>
Signed-off-by: Nick Hill <nhill@redhat.com>
Co-authored-by: Nick Hill <nhill@redhat.com>
This commit is contained in:
Michael Goin
2025-04-23 20:54:00 -06:00
committed by GitHub
parent 46e678bcff
commit ed50f46641
5 changed files with 75 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
vllm/usage/usage_lib.py
View File

@@ -19,6 +19,7 @@ import torch
import vllm.envs as envs
from vllm.connections import global_http_connection
from vllm.utils import cuda_device_count_stateless, cuda_get_device_properties
from vllm.version import __version__ as VLLM_VERSION
_config_home = envs.VLLM_CONFIG_ROOT
@@ -168,10 +169,9 @@ class UsageMessage:
# Platform information
from vllm.platforms import current_platform
if current_platform.is_cuda_alike():
device_property = torch.cuda.get_device_properties(0)
self.gpu_count = torch.cuda.device_count()
self.gpu_type = device_property.name
self.gpu_memory_per_device = device_property.total_memory
self.gpu_count = cuda_device_count_stateless()
self.gpu_type, self.gpu_memory_per_device = (
cuda_get_device_properties(0, ("name", "total_memory")))
if current_platform.is_cuda():
self.cuda_runtime = torch.version.cuda
self.provider = _detect_cloud_provider()