[Common] Help gc to collect unused model outputs #3643
Merged
Conversation
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
AlexanderDokuchaev
requested changes
Sep 2, 2025
| outputs = engine.infer(input_data) | ||
| processed_outputs = self._process_outputs(outputs) | ||
| self._register_statistics(processed_outputs, merged_statistics) | ||
| del outputs |
Collaborator
There was a problem hiding this comment.
Please add comment about reason to add del here.
AlexanderDokuchaev
approved these changes
Sep 16, 2025
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Outdated suggestions cannot be applied.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.
Changes
Variable which is containing model outputs during statistic collection is explicitly freed after each iteration
Reason for changes
The quantization cell was failing on my machine with 125GB of RAM in the SD v3 TorchFX notebook.
After checking that the pytorch code is not leaking by the pytorch profiler tool
I confirmed that the problem is not related to statistic collection directly. After that I forced the garbage collection by the
delstatement - and my memory didn't exceed healthy 50GB during the runtime, and SQ was successfully applied to the model (in contrast with previous runs)Related tickets
Tests
I'm not sure which test should I do with that