Replace multiprocessing DataLoader with thread-based embedding pipeline#41
Merged
Merged
Conversation
Image preprocessing is GIL-light as PIL/torchvision release the GIL, therefore a thread pool parallelizes it near-linearly without process spawn, pickling, or fork/spawn divergence... This commit introduce `shared/utils/image_pipeline.py`: thread-pool decode+preprocess overlapped with a main-thread batch forward. The embedding service implements device-aware concurrency - GPU = wide pool + intra-op threads pinned to 1 - CPU = small pool, forward keeps the cores This commit drops the hpc-inference git dependency entirely, unblocks PyPI release. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
egrace479
approved these changes
Jun 15, 2026
egrace479
left a comment
egrace479
left a comment
Member
There was a problem hiding this comment.
I can run it now (with more than 5 images).
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Image preprocessing is GIL-light as PIL/torchvision release the GIL, therefore a thread pool parallelizes it near-linearly without process spawn, pickling, or fork/spawn divergence...
This commit introduce
shared/utils/image_pipeline.py: thread-pool decode+preprocess overlapped with a main-thread batch forward.The embedding service implements device-aware concurrency
This commit drops the hpc-inference git dependency entirely, unblocks PyPI release.
Please see inline comment & docstrings for implementation details.