ROCm / pytorch

Tensors and Dynamic neural networks in Python with strong GPU acceleration
http://pytorch.org
Other
219 stars 50 forks source link

[release/2.2] Include ROCm patch version unconditionally in triton version #1410

Closed jithunnair-amd closed 2 months ago

jithunnair-amd commented 2 months ago

Follow-up to https://github.com/ROCm/pytorch/commit/16e830cee4530f2f97e4f45540400a158f270e86 to resolve issues with ROCm6.1.1 (and similar patch versions) where triton wheel wouldn't contain the patch version but torch wheel would have a dependency on the name with patch version.

Similar changes to https://github.com/ROCm/pytorch/commit/1c7726a612551daab5480f902e43ffc2cc400057, but makes patch version unconditional

This was requested by DC GPU team to simplify their scripts that use PyTorch wheels and determine the name using the ROCM build version shown here - http://mkdcvlnxapp01.amd.com/dashboard - under "Core/ROCm Release".

Tested via: http://ml-ci-internal.amd.com:8080/job/pytorch/job/dev/job/manylinux_rocm_wheels_test/56