[ROCm] Correct numerical issues in layer norm backwards kernel#140259
Closed
jataylo wants to merge 1 commit intopytorch:mainfrom
Closed
[ROCm] Correct numerical issues in layer norm backwards kernel#140259jataylo wants to merge 1 commit intopytorch:mainfrom
jataylo wants to merge 1 commit intopytorch:mainfrom
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/140259
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 496b18a with merge base 565a794 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
Contributor
|
@Mellonta has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
jianyuh
approved these changes
Nov 11, 2024
Collaborator
Author
|
@pytorchbot merge |
Collaborator
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
jataylo
added a commit
to ROCm/pytorch
that referenced
this pull request
Dec 4, 2024
…ch#140259) 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)
jataylo
added a commit
to ROCm/pytorch
that referenced
this pull request
Dec 4, 2024
…ch#140259) 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)
jataylo
added a commit
to ROCm/pytorch
that referenced
this pull request
Dec 4, 2024
…ch#140259) 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)
jataylo
added a commit
to ROCm/pytorch
that referenced
this pull request
Dec 4, 2024
…ch#140259) 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)
pobin6
pushed a commit
to pobin6/pytorch
that referenced
this pull request
Dec 5, 2024
…ch#140259) 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
pruthvistony
pushed a commit
to ROCm/pytorch
that referenced
this pull request
Dec 6, 2024
… kernel (pytorch#140259) (#1766) 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)
pruthvistony
pushed a commit
to ROCm/pytorch
that referenced
this pull request
Dec 6, 2024
… 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
pushed a commit
to ROCm/pytorch
that referenced
this pull request
Dec 6, 2024
… 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
pushed a commit
to ROCm/pytorch
that referenced
this pull request
Dec 13, 2024
… kernel (pytorch#140259) (#1766) 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)
rocm-mici
pushed a commit
to ROCm/pytorch
that referenced
this pull request
Dec 13, 2024
… kernel (pytorch#140259) (#1766) 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)
rocm-mici
pushed a commit
to ROCm/pytorch
that referenced
this pull request
Dec 13, 2024
… kernel (pytorch#140259) (#1766) 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)
rocm-mici
pushed a commit
to ROCm/pytorch
that referenced
this pull request
Dec 13, 2024
… 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
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
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
cuLoadWriteStridedInputswhich processes strided input and accumulates the partial gradients into shared memory.In this kernel (#87635) we truncated
meanandrstdfrom 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.
cc @jeffdaily @sunway513 @jithunnair-amd @pruthvistony @ROCmSupport @dllehr-amd @hongxiayang @naromero77amd