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, 05:30 UTC
[Misc] Define EP kernel arch list in Dockerfile (#25635)
Signed-off-by: Simon Mo <simon.mo@hey.com>
Simon Mo committed
Oct 6, 2025 at 17:05 UTC
8229280a9c74a2f01d486dee4165eba881b5579d
1 file changed
+1
-1
docker/Dockerfile
+1
-1
@@ -461,7 +461,7 @@ RUN set -eux; \
461
# Install EP kernels(pplx-kernels and DeepEP)
462
COPY tools/ep_kernels/install_python_libraries.sh install_python_libraries.sh
463
ENV CUDA_HOME=/usr/local/cuda
464
-RUN export TORCH_CUDA_ARCH_LIST="${TORCH_CUDA_ARCH_LIST:-9.0a+PTX}" \
464
+RUN export TORCH_CUDA_ARCH_LIST="${TORCH_CUDA_ARCH_LIST:-9.0a 10.0a+PTX}" \
465
&& bash install_python_libraries.sh
466
467
# CUDA image changed from /usr/local/nvidia to /usr/local/cuda in 12.8 but will