feat: Stability with fp16 for anima#2277
Merged
Merged
Conversation
… residual connections
|
Even after this PR is merged, it still causing NaN losses in fp16 precision. |
Owner
Author
Thank you for reporting. Does #2274 solve the issue? |
|
Yes, #2274 works fine without any problems. |
Owner
Author
|
Thank you, I will investigate. If I can't find the cause, I will consider adopting #2274 as is. |
Owner
Author
|
I've opened #2302, I hope this fixes the issue. |
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
2 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This pull request introduces an improvement to the
_forwardmethod inlibrary/anima_models.pyto enhance numerical stability during computations involving residual connections.Numerical stability improvements:
_forwardmethod, input tensorx_B_T_H_W_Dis now cast tofloat32if its original type isfloat16, ensuring better numerical stability in residual connections. Each module will handle casting back tofloat16as needed using an autocast context.Address the issue described in #2274.