Stop setting CUDA_VISIBLE_DEVICES internally in vLLM, add device_ids arg (#45026)

Signed-off-by: Tyler Michael Smith <tlrmchlsmth@gmail.com>
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Codex <codex@openai.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
Co-authored-by: kourosh hakhamaneshi <kouroshHakha@users.noreply.github.com>
This commit is contained in:
Tyler Michael Smith
2026-06-20 13:38:10 -07:00
committed by GitHub
co-authored by Claude Codex mergify[bot] <37929162+mergify[bot]@users.noreply.github.com> kourosh hakhamaneshi
parent e9de72fe6c
commit ebfbcfe46a
24 changed files with 722 additions and 185 deletions
+9
View File
@@ -685,6 +685,15 @@ class CudaPlatformBase(Platform):
# all the related functions work on real physical device ids.
# the major benefit of using NVML is that it will not initialize CUDA
class NvmlCudaPlatform(CudaPlatformBase):
@classmethod
@with_nvml_context
def device_control_id_to_physical_device_id(cls, device_id: str) -> int:
try:
return int(device_id)
except ValueError:
handle = pynvml.nvmlDeviceGetHandleByUUID(device_id)
return pynvml.nvmlDeviceGetIndex(handle)
@classmethod
@cache
@with_nvml_context