Skip to content

[BE] Install Helion without dependencies #156706

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Closed
atalman wants to merge 3 commits into from
Closed

[BE] Install Helion without dependencies #156706

atalman wants to merge 3 commits into from

Conversation

@atalman
Copy link
Contributor

@atalman atalman commented Jun 24, 2025 •
edited
Loading

After: #155513
Please see comment: #155513 (comment)

Here are the logs: https://github.com/pytorch/pytorch/actions/runs/15838529400/job/44646874281?pr=156664#step:6:16372

Looks like current workflow is :
Build triton - triton-3.4.0+git5389ed79-cp310-cp310-linux_x86_64.whl
Install Helion - Overwrite triton with production 3.3.1 and install production torch
Reinstall triton as final docker build step - triton-3.4.0+git5389ed79-cp310-cp310-linux_x86_64.whl

This makes it somewhat messy since we install both torch and triton from prod. This is something we want to avoid when building underlining docker images for CI

Log:

#55 311.4 + pip_install helion
#55 311.4 + as_jenkins conda run -n py_3.10 pip install --progress-bar off helion
#55 311.4 + sudo -E -H -u jenkins env -u SUDO_UID -u SUDO_GID -u SUDO_COMMAND -u SUDO_USER env PATH=/usr/local/nvidia/bin:/usr/local/cuda/bin:/opt/conda/envs/py_3.10/bin:/opt/conda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin LD_LIBRARY_PATH= conda run -n py_3.10 pip install --progress-bar off helion
#55 393.6 Collecting helion
#55 393.6   Downloading helion-0.0.7-py3-none-any.whl.metadata (14 kB)
#55 393.6 Collecting filecheck (from helion)
#55 393.6   Downloading filecheck-1.0.2-py3-none-any.whl.metadata (5.8 kB)
#55 393.6 Collecting torch>=2.7.0 (from helion)
#55 393.6   Downloading torch-2.7.1-cp310-cp310-manylinux_2_28_x86_64.whl.metadata (29 kB)
#55 393.6 Requirement already satisfied: typing-extensions>=4.0.0 in /opt/conda/envs/py_3.10/lib/python3.10/site-packages (from helion) (4.14.0)
#55 393.6 Requirement already satisfied: filelock in /opt/conda/envs/py_3.10/lib/python3.10/site-packages (from torch>=2.7.0->helion) (3.18.0)
#55 393.6 Requirement already satisfied: sympy>=1.13.3 in /opt/conda/envs/py_3.10/lib/python3.10/site-packages (from torch>=2.7.0->helion) (1.13.3)
#55 393.6 Requirement already satisfied: networkx in /opt/conda/envs/py_3.10/lib/python3.10/site-packages (from torch>=2.7.0->helion) (2.8.8)
#55 393.6 Requirement already satisfied: jinja2 in /opt/conda/envs/py_3.10/lib/python3.10/site-packages (from torch>=2.7.0->helion) (3.1.6)
#55 393.6 Requirement already satisfied: fsspec in /opt/conda/envs/py_3.10/lib/python3.10/site-packages (from torch>=2.7.0->helion) (2025.5.1)
#55 393.6 Collecting nvidia-cuda-nvrtc-cu12==12.6.77 (from torch>=2.7.0->helion)
#55 393.6   Downloading nvidia_cuda_nvrtc_cu12-12.6.77-py3-none-manylinux2014_x86_64.whl.metadata (1.5 kB)
#55 393.6 Collecting nvidia-cuda-runtime-cu12==12.6.77 (from torch>=2.7.0->helion)
#55 393.6   Downloading nvidia_cuda_runtime_cu12-12.6.77-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (1.5 kB)
#55 393.6 Collecting nvidia-cuda-cupti-cu12==12.6.80 (from torch>=2.7.0->helion)
#55 393.6   Downloading nvidia_cuda_cupti_cu12-12.6.80-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (1.6 kB)
#55 393.6 Collecting nvidia-cudnn-cu12==9.5.1.17 (from torch>=2.7.0->helion)
#55 393.6   Downloading nvidia_cudnn_cu12-9.5.1.17-py3-none-manylinux_2_28_x86_64.whl.metadata (1.6 kB)
#55 393.6 Collecting nvidia-cublas-cu12==12.6.4.1 (from torch>=2.7.0->helion)
#55 393.6   Downloading nvidia_cublas_cu12-12.6.4.1-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (1.5 kB)
#55 393.6 Collecting nvidia-cufft-cu12==11.3.0.4 (from torch>=2.7.0->helion)
#55 393.6   Downloading nvidia_cufft_cu12-11.3.0.4-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (1.5 kB)
#55 393.6 Collecting nvidia-curand-cu12==10.3.7.77 (from torch>=2.7.0->helion)
#55 393.6   Downloading nvidia_curand_cu12-10.3.7.77-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (1.5 kB)
#55 393.6 Collecting nvidia-cusolver-cu12==11.7.1.2 (from torch>=2.7.0->helion)
#55 393.6   Downloading nvidia_cusolver_cu12-11.7.1.2-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (1.6 kB)
#55 393.6 Collecting nvidia-cusparse-cu12==12.5.4.2 (from torch>=2.7.0->helion)
#55 393.6   Downloading nvidia_cusparse_cu12-12.5.4.2-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (1.6 kB)
#55 393.6 Collecting nvidia-cusparselt-cu12==0.6.3 (from torch>=2.7.0->helion)
#55 393.6   Downloading nvidia_cusparselt_cu12-0.6.3-py3-none-manylinux2014_x86_64.whl.metadata (6.8 kB)
#55 393.6 Collecting nvidia-nccl-cu12==2.26.2 (from torch>=2.7.0->helion)
#55 393.6   Downloading nvidia_nccl_cu12-2.26.2-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (2.0 kB)
#55 393.6 Collecting nvidia-nvtx-cu12==12.6.77 (from torch>=2.7.0->helion)
#55 393.6   Downloading nvidia_nvtx_cu12-12.6.77-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (1.6 kB)
#55 393.6 Collecting nvidia-nvjitlink-cu12==12.6.85 (from torch>=2.7.0->helion)
#55 393.6   Downloading nvidia_nvjitlink_cu12-12.6.85-py3-none-manylinux2010_x86_64.manylinux_2_12_x86_64.whl.metadata (1.5 kB)
#55 393.6 Collecting nvidia-cufile-cu12==1.11.1.6 (from torch>=2.7.0->helion)
#55 393.6   Downloading nvidia_cufile_cu12-1.11.1.6-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (1.5 kB)
#55 393.6 Collecting triton==3.3.1 (from torch>=2.7.0->helion)
#55 393.6   Downloading triton-3.3.1-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.metadata (1.5 kB)
#55 393.6 Requirement already satisfied: setuptools>=40.8.0 in /opt/conda/envs/py_3.10/lib/python3.10/site-packages (from triton==3.3.1->torch>=2.7.0->helion) (80.9.0)
#55 393.6 Requirement already satisfied: mpmath<1.4,>=1.1.0 in /opt/conda/envs/py_3.10/lib/python3.10/site-packages (from sympy>=1.13.3->torch>=2.7.0->helion) (1.3.0)
#55 393.6 Requirement already satisfied: MarkupSafe>=2.0 in /opt/conda/envs/py_3.10/lib/python3.10/site-packages (from jinja2->torch>=2.7.0->helion) (3.0.2)
#55 393.6 Downloading helion-0.0.7-py3-none-any.whl (149 kB)
#55 393.6 Downloading torch-2.7.1-cp310-cp310-manylinux_2_28_x86_64.whl (821.2 MB)
#55 393.6 Downloading nvidia_cublas_cu12-12.6.4.1-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (393.1 MB)
#55 393.6 Downloading nvidia_cuda_cupti_cu12-12.6.80-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (8.9 MB)
#55 393.6 Downloading nvidia_cuda_nvrtc_cu12-12.6.77-py3-none-manylinux2014_x86_64.whl (23.7 MB)
#55 393.6 Downloading nvidia_cuda_runtime_cu12-12.6.77-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (897 kB)
#55 393.6 Downloading nvidia_cudnn_cu12-9.5.1.17-py3-none-manylinux_2_28_x86_64.whl (571.0 MB)
#55 393.6 Downloading nvidia_cufft_cu12-11.3.0.4-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (200.2 MB)
#55 393.6 Downloading nvidia_cufile_cu12-1.11.1.6-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (1.1 MB)
#55 393.6 Downloading nvidia_curand_cu12-10.3.7.77-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (56.3 MB)
#55 393.6 Downloading nvidia_cusolver_cu12-11.7.1.2-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (158.2 MB)
#55 393.6 Downloading nvidia_cusparse_cu12-12.5.4.2-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (216.6 MB)
#55 393.6 Downloading nvidia_cusparselt_cu12-0.6.3-py3-none-manylinux2014_x86_64.whl (156.8 MB)
#55 393.6 Downloading nvidia_nccl_cu12-2.26.2-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (201.3 MB)
#55 393.6 Downloading nvidia_nvjitlink_cu12-12.6.85-py3-none-manylinux2010_x86_64.manylinux_2_12_x86_64.whl (19.7 MB)
#55 393.6 Downloading nvidia_nvtx_cu12-12.6.77-py3-none-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (89 kB)
#55 393.6 Downloading triton-3.3.1-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (155.6 MB)
#55 393.6 Downloading filecheck-1.0.2-py3-none-any.whl (23 kB)
#55 393.6 Installing collected packages: nvidia-cusparselt-cu12, triton, nvidia-nvtx-cu12, nvidia-nvjitlink-cu12, nvidia-nccl-cu12, nvidia-curand-cu12, nvidia-cufile-cu12, nvidia-cuda-runtime-cu12, nvidia-cuda-nvrtc-cu12, nvidia-cuda-cupti-cu12, nvidia-cublas-cu12, filecheck, nvidia-cusparse-cu12, nvidia-cufft-cu12, nvidia-cudnn-cu12, nvidia-cusolver-cu12, torch, helion
#55 393.6   Attempting uninstall: triton
#55 393.6     Found existing installation: triton 3.4.0+git5389ed79
#55 393.6     Uninstalling triton-3.4.0+git5389ed79:
#55 393.6       Successfully uninstalled triton-3.4.0+git5389ed79
#55 393.6 Successfully installed filecheck-1.0.2 helion-0.0.7 nvidia-cublas-cu12-12.6.4.1 nvidia-cuda-cupti-cu12-12.6.80 nvidia-cuda-nvrtc-cu12-12.6.77 nvidia-cuda-runtime-cu12-12.6.77 nvidia-cudnn-cu12-9.5.1.17 nvidia-cufft-cu12-11.3.0.4 nvidia-cufile-cu12-1.11.1.6 nvidia-curand-cu12-10.3.7.77 nvidia-cusolver-cu12-11.7.1.2 nvidia-cusparse-cu12-12.5.4.2 nvidia-cusparselt-cu12-0.6.3 nvidia-nccl-cu12-2.26.2 nvidia-nvjitlink-cu12-12.6.85 nvidia-nvtx-cu12-12.6.77 torch-2.7.1 triton-3.3.1
#55 393.6 
#55 DONE 428.8s

#56 [final  1/30] COPY --from=triton-builder /opt/triton /opt/triton
#56 DONE 0.0s

#57 [final  2/30] RUN if [ -n "yes" ] || [ -n "" ]; then pip install /opt/triton/*.whl; chown -R jenkins:jenkins /opt/conda; fi
#57 0.823 Processing /opt/triton/triton-3.4.0+git5389ed79-cp310-cp310-linux_x86_64.whl
#57 2.263 Requirement already satisfied: setuptools>=40.8.0 in /opt/conda/envs/py_3.10/lib/python3.10/site-packages (from triton==3.4.0+git5389ed79) (80.9.0)
#57 2.589 Installing collected packages: triton
#57 6.405 Successfully installed triton-3.4.0+git5389ed79
#57 6.405 WARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager, possibly rendering your system unusable. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv. Use the --root-user-action option if you know what you are doing and want to suppress this warning.
#57 DONE 86.5s

@atalman atalman requested a review from jeffdaily as a code owner June 24, 2025 15:26
@pytorch-bot
Copy link

pytorch-bot bot commented Jun 24, 2025 •
edited
Loading

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/156706

Note: Links to docs will display an error until the docs builds have been completed.

✅ You can merge normally! (1 Unrelated Failure)

As of commit be8ba18 with merge base 795a6a0 (image):

UNSTABLE - The following job is marked as unstable, possibly due to flakiness on trunk:

This comment was automatically generated by Dr. CI and updates every 15 minutes.

@pytorch-bot pytorch-bot bot added the topic: not user facing topic category label Jun 24, 2025
@atalman atalman changed the title Install Helion from before building triton [BE] Install Helion from before building triton Jun 24, 2025
@atalman atalman changed the title [BE] Install Helion from before building triton [BE] Install Helion before building triton Jun 24, 2025
@atalman atalman changed the title [BE] Install Helion before building triton [BE] Install Helion without dependencies Jun 24, 2025
malfet
malfet approved these changes Jun 24, 2025
View reviewed changes
Copy link
Contributor

@malfet malfet left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@atalman
Copy link
Contributor Author

atalman commented Jun 25, 2025

@pytorchmergebot merge -i

@pytorch-bot pytorch-bot bot added the ciflow/trunk Trigger trunk jobs on your pull request label Jun 25, 2025
@pytorchmergebot
Copy link
Collaborator

pytorchmergebot commented Jun 25, 2025

Merge started

Your change will be merged while ignoring the following 1 checks: pull / cuda12.8-py3.10-gcc9-sm75 / test (pr_time_benchmarks, 1, 1, linux.g4dn.metal.nvidia.gpu, unstable)

Learn more about merging in the wiki.

Questions? Feedback? Please reach out to the PyTorch DevX Team

Advanced Debugging
Check the merge workflow status
here

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@malfet malfet malfet approved these changes

@oulgen oulgen oulgen approved these changes

@jeffdaily jeffdaily Awaiting requested review from jeffdaily jeffdaily is a code owner

Assignees

No one assigned

Labels

ciflow/trunk Trigger trunk jobs on your pull request Merged topic: not user facing topic category

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

4 participants

@atalman @pytorchmergebot @malfet @oulgen