facebookresearch/faiss
1
0
Fork 0
mirror of https://github.com/facebookresearch/faiss.git synced 2026-10-11 22:50:00 +00:00
Code Issues Packages Projects Releases Wiki Activity

Upgrade cuVS Version to 26.02 (#4945)

Browse Source 
Summary:
Pull Request resolved: https://github.com/facebookresearch/faiss/pull/4945

Pull Request resolved: https://github.com/facebookresearch/faiss/pull/4788

Reviewed By: yingufan

Differential Revision: D96931953

Pulled By: mnorris11

fbshipit-source-id: 7d7a9b563048dc10bd44d22035216c7aa98b0003
This commit is contained in:
Tarang Jain
2026-03-17 21:58:50 -07:00
committed by meta-codesync[bot]
parent 0f7e65e53d
commit 02d609caff
16 changed files with 145 additions and 72 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.github/actions/build_cmake/action.yml
+2 -2
View File
@@ -74,7 +74,7 @@ runs:
conda install -y -q cuda-toolkit=12.6 gxx_linux-64=12.4 -c "nvidia/label/cuda-12.6"
# and CUDA from cuVS channel for cuVS builds
elif [ "${{ inputs.cuvs }}" = "ON" ]; then
conda install -y -q libcuvs=25.10 'cuda-version=12.6' cuda-toolkit=12.6 gxx_linux-64=13.4.0 -c rapidsai -c rapidsai-nightly -c conda-forge
conda install -y -q libcuvs=26.02 'cuda-version=12.9' cuda-toolkit=12.9 sysroot_linux-64=2.34 -c rapidsai -c rapidsai-nightly -c conda-forge
fi
# install SVS runtime for SVS builds
@@ -87,7 +87,7 @@ runs:
: # skip torch install via conda, we need to install via pip to get
# ROCm-enabled version until it's supported in conda by PyTorch
elif [ "${{ inputs.gpu }}" = "ON" ]; then
conda install -y -q "pytorch>=2.7" "pytorch-gpu>=2.7" -c pytorch -c "nvidia/label/12.6"
conda install -y -q "pytorch>=2.7" "pytorch-gpu>=2.7" -c pytorch -c "nvidia/label/12.9"
else
# TestLowLevelIVF.IVFRQ hangs on pytorch>=2.7, so left it as <2.5 for now.
conda install -y -q "pytorch<2.5" -c pytorch