Skip to content
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

[AUTOGENERATED] [release/2.5] Cherry-pick PR-1767 #1794

Draft
wants to merge 1 commit into
base: release/2.5
Choose a base branch
from
Draft

[AUTOGENERATED] [release/2.5] Cherry-pick PR-1767 #1794

wants to merge 1 commit into from
+2 −2

Conversation

rocm-mici
Copy link

Cherry-pick of #1767

@jataylo
[release/2.2] [ROCm] Correct numerical issues in layer norm backwards…
da78ffb 
… kernel (pytorch#140259) (#1767)

It was raised that the backwards layer norm on AMD was slightly off the
accuracy of the equivalent NVIDIA implementation.

On AMD we call into a helper kernel `cuLoadWriteStridedInputs` which
processes strided input and accumulates the partial gradients into
shared memory.

In this kernel (pytorch#87635) we
truncated `mean` and `rstd` from T_ACC type to T which causes numerical
issues in the warp buffers created in this kernel. This PR will use the
correct accumulator type for mean and rstd.

Note: Only AMD call into this call stack for backwards layer norm, so
this was not an issue for NV.

Pull Request resolved: pytorch#140259
Approved by: https://github.com/jianyuh

(cherry picked from commit 001f736)

Fixes #ISSUE_NUMBER
@rocm-mici rocm-mici mentioned this pull request Dec 13, 2024
Merged
@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Dec 13, 2024 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as FAILURE
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Dec 17, 2024 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as FAILURE
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Dec 18, 2024 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as FAILURE
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Dec 19, 2024 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as FAILURE
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Dec 23, 2024 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as FAILURE
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Dec 26, 2024 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as FAILURE
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Jan 6, 2025 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as FAILURE
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Jan 8, 2025 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as FAILURE
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Jan 8, 2025 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as FAILURE
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Jan 10, 2025 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as FAILURE
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Jan 14, 2025 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as ABORTED
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Jan 15, 2025 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as ABORTED
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Jan 15, 2025 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as ABORTED
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Jan 16, 2025 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as FAILURE
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Jan 16, 2025 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as ABORTED
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

rocm-repo-management-api bot commented Jan 16, 2025 •
edited
Loading

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit finished as FAILURE
Links: Blue Ocean view / Build artifacts

@rocm-repo-management-api
Copy link

Jenkins build for da78ffb5ae9c5d8d1b745237839e04bd4a1fd01b commit is in progress
Links: Blue Ocean view / Build artifacts

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@rocm-mici @jataylo