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

Convert formatting to use ruff instead of yapf + isort (#26247)

Browse Source 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-10-05 15:06:22 +01:00
committed by GitHub
parent 17edd8a807
commit d6953beb91
1508 changed files with 115244 additions and 94146 deletions
Download Patch File Download Diff File Expand all files Collapse all files

76
vllm/usage/usage_lib.py
View File

@@ -68,8 +68,7 @@ def is_usage_stats_enabled():
no_usage_stats = envs.VLLM_NO_USAGE_STATS
do_not_track_file = os.path.exists(_USAGE_STATS_DO_NOT_TRACK_PATH)
_USAGE_STATS_ENABLED = not (do_not_track or no_usage_stats
or do_not_track_file)
_USAGE_STATS_ENABLED = not (do_not_track or no_usage_stats or do_not_track_file)
return _USAGE_STATS_ENABLED
@@ -80,9 +79,11 @@ def _get_current_timestamp_ns() -> int:
def _detect_cloud_provider() -> str:
# Try detecting through vendor file
vendor_files = [
"/sys/class/dmi/id/product_version", "/sys/class/dmi/id/bios_vendor",
"/sys/class/dmi/id/product_version",
"/sys/class/dmi/id/bios_vendor",
"/sys/class/dmi/id/product_name",
"/sys/class/dmi/id/chassis_asset_tag", "/sys/class/dmi/id/sys_vendor"
"/sys/class/dmi/id/chassis_asset_tag",
"/sys/class/dmi/id/sys_vendor",
]
# Mapping of identifiable strings to cloud providers
cloud_identifiers = {
@@ -152,39 +153,53 @@ class UsageMessage:
self.log_time: Optional[int] = None
self.source: Optional[str] = None
def report_usage(self,
model_architecture: str,
usage_context: UsageContext,
extra_kvs: Optional[dict[str, Any]] = None) -> None:
t = Thread(target=self._report_usage_worker,
args=(model_architecture, usage_context, extra_kvs or {}),
daemon=True)
def report_usage(
self,
model_architecture: str,
usage_context: UsageContext,
extra_kvs: Optional[dict[str, Any]] = None,
) -> None:
t = Thread(
target=self._report_usage_worker,
args=(model_architecture, usage_context, extra_kvs or {}),
daemon=True,
)
t.start()
def _report_usage_worker(self, model_architecture: str,
usage_context: UsageContext,
extra_kvs: dict[str, Any]) -> None:
def _report_usage_worker(
self,
model_architecture: str,
usage_context: UsageContext,
extra_kvs: dict[str, Any],
) -> None:
self._report_usage_once(model_architecture, usage_context, extra_kvs)
self._report_continuous_usage()
def _report_usage_once(self, model_architecture: str,
usage_context: UsageContext,
extra_kvs: dict[str, Any]) -> None:
def _report_usage_once(
self,
model_architecture: str,
usage_context: UsageContext,
extra_kvs: dict[str, Any],
) -> None:
# Platform information
from vllm.platforms import current_platform
if current_platform.is_cuda_alike():
self.gpu_count = cuda_device_count_stateless()
self.gpu_type, self.gpu_memory_per_device = (
cuda_get_device_properties(0, ("name", "total_memory")))
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
if current_platform.is_tpu():
try:
import torch_xla
self.gpu_count = torch_xla.runtime.world_size()
self.gpu_type = torch_xla.tpu.get_tpu_type()
self.gpu_memory_per_device = (
torch_xla.core.xla_model.get_memory_info()["bytes_limit"])
self.gpu_memory_per_device = torch_xla.core.xla_model.get_memory_info()[
"bytes_limit"
]
except Exception:
logger.exception("Failed to collect TPU information")
self.provider = _detect_cloud_provider()
@@ -195,11 +210,13 @@ class UsageMessage:
info = cpuinfo.get_cpu_info()
self.num_cpu = info.get("count", None)
self.cpu_type = info.get("brand_raw", "")
self.cpu_family_model_stepping = ",".join([
str(info.get("family", "")),
str(info.get("model", "")),
str(info.get("stepping", ""))
])
self.cpu_family_model_stepping = ",".join(
[
str(info.get("family", "")),
str(info.get("model", "")),
str(info.get("stepping", "")),
]
)
# vLLM information
self.context = usage_context.value
@@ -207,10 +224,9 @@ class UsageMessage:
self.model_architecture = model_architecture
# Environment variables
self.env_var_json = json.dumps({
env_var: getattr(envs, env_var)
for env_var in _USAGE_ENV_VARS_TO_COLLECT
})
self.env_var_json = json.dumps(
{env_var: getattr(envs, env_var) for env_var in _USAGE_ENV_VARS_TO_COLLECT}
)
# Metadata
self.log_time = _get_current_timestamp_ns()