[ROCm][CI] Run PR-Based workflow runs on mi300 nodes.#167225
[ROCm][CI] Run PR-Based workflow runs on mi300 nodes.#167225amdfaa wants to merge 9 commits intopytorch:mainfrom
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/167225
Note: Links to docs will display an error until the docs builds have been completed. ❌ 1 New Failure, 1 Unrelated FailureAs of commit 326f83c with merge base 9b4ac45 ( BROKEN TRUNK - The following job failed but were present on the merge base:👉 Rebase onto the `viable/strict` branch to avoid these failures
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
daab6e9 to
326f83c
Compare
|
@pytorchbot merge -f "Lint and other failures seem unrelated; PR labels triggered MI300 workflows successfully" |
Merge startedYour change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Please use Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
Fixes issue with uploading artifacts, which was inadvertently disabled for some renamed workflows via #167225 Pull Request resolved: #167483 Approved by: https://github.com/jeffdaily
This PR is meant to swap the PR-based ciflow tags from the mi200 nodes (less stable) to the mi300 nodes (more stable). This will ensure that developers see consistent testing on their PRs as well as on main. This PR does all of the following: - Rename rocm.yml to rocm-mi200.yml : for clarity - Add ciflow/rocm-mi200 trigger to rocm-mi200.yml : for devs who want to opt-in to single-GPU unit tests on MI200 - Move ciflow/rocm trigger from rocm-mi200.yml to rocm-mi300.yml : so PRs target MI300 runners by default - Rename inductor-rocm.yml to inductor-rocm-mi200.yml : for clarity - Remove ciflow/inductor-rocm trigger from inductor-rocm-mi200.yml : prevent MI200 inductor config unit tests being triggered by default - Add ciflow/inductor-rocm-mi200 trigger to inductor-rocm-mi200.yml : for devs who want to opt-in to inductor config unit tests on MI200 - Move ciflow/periodic trigger from periodic-rocm-mi200.yml to periodic-rocm-mi300.yml : so PRs target MI300 runners by default Pull Request resolved: pytorch#167225 Approved by: https://github.com/jeffdaily, https://github.com/huydhn Co-authored-by: Jithun Nair <jithun.nair@amd.com>
) Fixes issue with uploading artifacts, which was inadvertently disabled for some renamed workflows via pytorch#167225 Pull Request resolved: pytorch#167483 Approved by: https://github.com/jeffdaily
This PR is meant to swap the PR-based ciflow tags from the mi200 nodes (less stable) to the mi300 nodes (more stable). This will ensure that developers see consistent testing on their PRs as well as on main. This PR does all of the following:
Rename rocm.yml to rocm-mi200.yml : for clarity
Add ciflow/rocm-mi200 trigger to rocm-mi200.yml : for devs who want to opt-in to single-GPU unit tests on MI200
Move ciflow/rocm trigger from rocm-mi200.yml to rocm-mi300.yml : so PRs target MI300 runners by default
Rename inductor-rocm.yml to inductor-rocm-mi200.yml : for clarity
Remove ciflow/inductor-rocm trigger from inductor-rocm-mi200.yml : prevent MI200 inductor config unit tests being triggered by default
Add ciflow/inductor-rocm-mi200 trigger to inductor-rocm-mi200.yml : for devs who want to opt-in to inductor config unit tests on MI200
Move ciflow/periodic trigger from periodic-rocm-mi200.yml to periodic-rocm-mi300.yml : so PRs target MI300 runners by default
cc @jeffdaily @sunway513 @jithunnair-amd @pruthvistony @ROCmSupport @dllehr-amd @jataylo @hongxiayang @naromero77amd