Removes unnecessary cloning #6761
Merged
Conversation
This file contains 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
About clone_tensors_for_torch_save() function: if the `item.device` is different from `device` input, clone() is not required because to() function do the same thing. * With the unnecessary clone(), I observed memory bloat. After I removed the unnecessary clone(), it's solved.
tjruwase
approved these changes
Nov 20, 2024
|
@swigls, thanks! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
clone_tensors_for_torch_save()function:When the
item.deviceis different fromdeviceinput,tensor.clone()is not actually required becauseto()function also clones the original tensor.+) I observed memory bloat under following conditions:
transformersframework withZeRO-0andZeRO-1configuration.clone_tensors_for_torch_save()After I removed the unnecessary
clone(), seems like the problem is solved.