Add collate_fn parameter support to DataLoader #48
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
- Add collate_fn parameter to core DataLoader class for custom batch collation - Implement collate_fn support across all backends (JAX, PyTorch, TensorFlow) - JAX backend applies collate_fn to batched data from dataset indexing - PyTorch backend uses custom collate_fn or defaults to _numpy_collate - TensorFlow backend uses tf.data.Dataset.map() for collate_fn application - Add Callable import to typing imports for proper type annotations - Add comprehensive test_collate_fn function in tests.ipynb - Update BaseDataLoader interface to include collate_fn parameter - Add documentation example in core.ipynb demonstrating usage The collate_fn parameter behaves consistently with PyTorch's DataLoader, allowing users to customize how individual samples are combined into batches.
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
- Fix test_collate_fn to properly handle PyTorch list format vs JAX batched format - Add support for HuggingFace datasets with PyTorch backend (list of dicts) - Ensure collate_fn works correctly across all backends (JAX, PyTorch, TensorFlow) - All collate_fn tests now pass for available backends
- Add tf_collate_wrapper to handle TensorFlow's argument unpacking behavior - TensorFlow map() calls functions with unpacked args (features, labels) - Wrapper packs them back into tuple format expected by collate_fn - Ensures result is properly unpacked for TensorFlow consumption - Fixes TypeError in CI when TensorFlow is installed
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.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 fixes #38
Summary
collate_fnparameter toDataLoaderfor custom batch collation functionalityChanges
collate_fnparameter with proper type annotationscollate_fnto batched data from dataset indexingcollate_fnor defaults to_numpy_collatetf.data.Dataset.map()forcollate_fnapplicationCallableimport to typing importstest_collate_fnfunction intests.ipynbwith cross-backend validationcore.ipynbBehavior
The
collate_fnparameter behaves consistently with PyTorch's DataLoader, allowing users to customize how individual samples are combined into batches. Whencollate_fn=None, each backend uses its default collation behavior.Test plan
collate_fnfunctionality with JAX backendcollate_fnfunctionality with PyTorch backendcollate_fnfunctionality with TensorFlow backendcollate_fn=Noneuses default behaviornbdev_testto ensure all tests pass