-
Notifications
You must be signed in to change notification settings - Fork 25.7k
Add HIP to the memory profiler device list #38795
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.
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
Conversation
Summary: Add HIP alongside CUDA Test Plan: rocm CI Reviewers: Subscribers: Tasks: Tags: [ghstack-poisoned]
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
let's wait for CI
💊 CI failures summary and remediationsAs of commit e79700f (more details on the Dr. CI page):
ci.pytorch.org: 1 failedThis comment was automatically generated by Dr. CI (expand for details).Follow this link to opt-out of these comments for your Pull Requests.Please report bugs/suggestions on the GitHub issue tracker or post in the (internal) Dr. CI Users group. This comment has been revised 8 times. |
|
Beat me to it. I was just about to push the same PR. I tested this PR locally and it passed. BTW ROCm CI might be failing now due to an unrelated change that landed while this memory profiler failure was blocking ROCm CI. |
|
I remember rocm build was not passing before my PR landed |
Summary: Add HIP alongside CUDA Test Plan: rocm CI Differential Revision: [D21665627](https://our.internmc.facebook.com/intern/diff/D21665627) [ghstack-poisoned]
Summary: Add HIP alongside CUDA Test Plan: rocm CI Differential Revision: [D21665627](https://our.internmc.facebook.com/intern/diff/D21665627) [ghstack-poisoned]
|
"test_memory_profiler (main.TestAutograd) ... ok" |
|
and bloat16 tests are broken on the trunk |
|
@ilia-cher merged this pull request in 83fa3f1. |
|
bfloat16 tests should be fixed by this #38823, we were just waiting for it to merge. We are aware of the build dashboard, and check it if not once but many times a day, due to upstream commits frequently landing without ROCm CI passing. We need help from the community and the maintainers to make ROCm CI more stable, e.g. not landing PRs with known breakage for ROCm back-end. Meanwhile, we are trying our best to improve the CI maintenance process and our software quality. |
Stack from ghstack:
Summary:
Add HIP alongside CUDA
Test Plan:
rocm CI
Differential Revision: D21665627