[backport] Update CUDA docker image and NCCL. (#8139) (#8162)

* Update CUDA docker image and NCCL. (#8139)

* Rest of the CI.

* CPU test dependencies.
This commit is contained in:
Jiaming Yuan
2022-08-12 18:57:42 +08:00
committed by GitHub
parent a55d3bdde2
commit 39c1488a42
13 changed files with 68 additions and 45 deletions

View File

@@ -4,7 +4,6 @@ ARG CUDA_VERSION_ARG
# Install all basic requirements
RUN \
rpm --erase gpg-pubkey-7fa2af80* && \
curl -fsSL https://developer.download.nvidia.com/compute/cuda/repos/rhel7/x86_64/D42D0685.pub | sed '/^Version/d' \
> /etc/pki/rpm-gpg/RPM-GPG-KEY-NVIDIA && \
yum install -y epel-release centos-release-scl && \
@@ -25,12 +24,10 @@ RUN \
# NCCL2 (License: https://docs.nvidia.com/deeplearning/sdk/nccl-sla/index.html)
RUN \
export CUDA_SHORT=`echo $CUDA_VERSION_ARG | grep -o -E '[0-9]+\.[0-9]'` && \
export NCCL_VERSION=2.8.3-1 && \
wget -nv -nc https://developer.download.nvidia.com/compute/machine-learning/repos/rhel7/x86_64/nvidia-machine-learning-repo-rhel7-1.0.0-1.x86_64.rpm && \
rpm -i nvidia-machine-learning-repo-rhel7-1.0.0-1.x86_64.rpm && \
export NCCL_VERSION=2.13.4-1 && \
yum-config-manager --add-repo http://developer.download.nvidia.com/compute/cuda/repos/rhel7/x86_64/cuda-rhel7.repo && \
yum -y update && \
yum install -y libnccl-${NCCL_VERSION}+cuda${CUDA_SHORT} libnccl-devel-${NCCL_VERSION}+cuda${CUDA_SHORT} libnccl-static-${NCCL_VERSION}+cuda${CUDA_SHORT} && \
rm -f nvidia-machine-learning-repo-rhel7-1.0.0-1.x86_64.rpm;
yum install -y libnccl-${NCCL_VERSION}+cuda${CUDA_SHORT} libnccl-devel-${NCCL_VERSION}+cuda${CUDA_SHORT} libnccl-static-${NCCL_VERSION}+cuda${CUDA_SHORT}
ENV PATH=/opt/python/bin:/opt/maven/bin:$PATH
ENV CC=/opt/rh/devtoolset-8/root/usr/bin/gcc