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 4, 15:00 UTC
[Bugfix] Install libcublas-dev in Dockerfile for FlashInfer CuTe DSL JIT (#39855)
Signed-off-by: esmeetu <jasonailu87@gmail.com> Co-authored-by: Roger Wang <hey@rogerw.io>
Roy Wang committed
Apr 27, 2026 at 15:47 UTC
985961345a13f3e3bb15d29c94b011ba9a6b858b
1 file changed
+1
-1
docker/Dockerfile
+1
-1
@@ -538,7 +538,7 @@ RUN CUDA_VERSION_DASH=$(echo $CUDA_VERSION | cut -d. -f1,2 | tr '.' '-') && \
538
cuda-nvrtc-${CUDA_VERSION_DASH} \
539
cuda-cuobjdump-${CUDA_VERSION_DASH} \
540
libcurand-dev-${CUDA_VERSION_DASH} \
541
- libcublas-${CUDA_VERSION_DASH} \
541
+ libcublas-dev-${CUDA_VERSION_DASH} \
542
# Required by fastsafetensors (fixes #20384)
543
libnuma-dev && \
544
# Fixes nccl_allocator requiring nccl.h at runtime