A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 14, 22:35 UTC
[Misc] `VLLM_TARGET_DEVICE.lower()` (#22101)
Signed-off-by: NickLucche <nlucches@redhat.com>
Nicolò Lucchesi committed
Aug 2, 2025 at 04:41 UTC
57393715e804387588241fbdb4ec94a7570230b6
1 file changed
+1
-1
vllm/envs.py
+1
-1
@@ -213,7 +213,7 @@ environment_variables: dict[str, Callable[[], Any]] = {
213
# Target device of vLLM, supporting [cuda (by default),
214
# rocm, neuron, cpu]
215
"VLLM_TARGET_DEVICE":
216
- lambda: os.getenv("VLLM_TARGET_DEVICE", "cuda"),
216
+ lambda: os.getenv("VLLM_TARGET_DEVICE", "cuda").lower(),
217
218
# Maximum number of compilation jobs to run in parallel.
219
# By default this is the number of CPUs