Hi community, I have built pt2.0 with gpu. The code I used is from github repo.
I used python setup.py install and when run python -m torch.utils.collect_env I cannot find triton.
Hence I tried to build triton with .github/scripts/build_triton_wheel.py.
However during the process it is downloading cuda 12.0.
I know from previous post that cuda 11.7 is prefered, why is there a difference on cuda version for code in the same repo?
The build works for me and creates a wheel without downloading any CUDA version.
Hi Sir ptrblck.
- is it possible to share how you build triton in pytorch directory? I want to double check to make sure my build cmd is legit.
- I thought build pytorch with python setup.py develop will enable the build of triton by default, but it turned out not the case, I am wondering how to make sure triton is enabled in my pytorch build.
- I used pytorch 2.0-rc2 and the pined triton commit is d54c04abe2c3e67b2139c68cdbda87b59e8dd01b I think there do exist a different cuda requirement since pytorch 2.0 stables at 11.7 and it seems that triton downloads 12.0.
- I used
python .github/scripts/build_triton_wheel.py --py-version 3.8
which createdpytorch_triton-2.0.0+b8b470bc59-cp38-cp38-linux_x86_64.whl
and which matches the pinned commit in.github/ci_commit_pins/triton.txt
. - I think you are correct and you would need to build
pytorch-triton
as a dependency as seen in my command. - Yes, RC2 used
d54c04abe2c3e67b2139c68cdbda87b59e8dd01b
which was updated tob8b470bc59
recently. I don’t think Triton should and is downloading any CUDA version, but ships withptxas
and its ownlibdevice.10.bc
library. I also did not see any CUDA download during my build and would not even know how it could install a full CUDA toolkit.
1 Like