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 12, 04:45 UTC
[CI/Build] Fixes missing runtime dependencies (#29822)
Signed-off-by: bbartels <benjamin@bartels.dev>
Benjamin Bartels committed
Dec 2, 2025 at 18:21 UTC
2d613de9aef3ef25e9adc8887ac0388da092b500
1 file changed
+6
-1
docker/Dockerfile
+6
-1
@@ -364,7 +364,12 @@ RUN CUDA_VERSION_DASH=$(echo $CUDA_VERSION | cut -d. -f1,2 | tr '.' '-') && \
364
cuda-cudart-${CUDA_VERSION_DASH} \
365
cuda-nvrtc-${CUDA_VERSION_DASH} \
366
cuda-cuobjdump-${CUDA_VERSION_DASH} \
367
- libcublas-${CUDA_VERSION_DASH} && \
367
+ # https://github.com/vllm-project/vllm/issues/29590
368
+ libcurand-dev-${CUDA_VERSION_DASH} \
369
+ libcublas-${CUDA_VERSION_DASH} \
370
+ # Fixes nccl_allocator requiring nccl.h at runtime
371
+ # https://github.com/vllm-project/vllm/blob/1336a1ea244fa8bfd7e72751cabbdb5b68a0c11a/vllm/distributed/device_communicators/pynccl_allocator.py#L22
372
+ libnccl-dev && \
373
rm -rf /var/lib/apt/lists/*
374
375
ARG PIP_INDEX_URL UV_INDEX_URL