fix: PIL Image Memory Leaks in Dataset Builders #194
+505
−293
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
Contributor
|
✅ DCO Check Passed Thanks @samiuc, all your commits are properly signed off. 🎉 |
Merge ProtectionsYour pull request matches the following merge protections and will not be merged until they are valid. 🔴 Require two reviewer for test updatesThis rule is failing.When test data is updated, we require two reviewers
🟢 Enforce conventional commitWonderful, this rule succeeded.Make sure that we follow https://www.conventionalcommits.org/en/v1.0.0/
|
samiuc
changed the title
I, samiuc <[email protected]>, hereby add my Signed-off-by to this commit: a592458 I, samiuc <[email protected]>, hereby add my Signed-off-by to this commit: ac622cc I, samiuc <[email protected]>, hereby add my Signed-off-by to this commit: 3648bd9 Signed-off-by: samiuc <[email protected]>
I, samiuc <[email protected]>, hereby add my Signed-off-by to this commit: a592458 I, samiuc <[email protected]>, hereby add my Signed-off-by to this commit: ac622cc I, samiuc <[email protected]>, hereby add my Signed-off-by to this commit: 3648bd9 I, samiuc <[email protected]>, hereby add my Signed-off-by to this commit: c05b85e Signed-off-by: samiuc <[email protected]>
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Outdated suggestions cannot be applied.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.
During our testing, we found that the memory grows unbounded when processing large datasets, eventually causing OOM errors caused by the PIL Image objects opened during dataset iteration are never closed. Images opened via Image.open() in builders and decoded via Features_Image().decode_example() when loading from parquet accumulate in memory.
So the code changes in this PR add an auto-close PIL Images in as_record_dict() after serialization. Added _close_images()
method to DatasetRecord and DatasetRecordWithPrediction that releases image resources. Also added try/finally blocks in builders (pixparse, funsd, xfund, file_dataset) to close images after local processing.