Fix confusing weight loading logging for legacy models #439
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Outdated suggestions cannot be applied.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.
Summary
.h5filesProblem
When loading legacy TensorFlow/Keras weights into separate PyTorch components (backbone vs head), users saw:
Solution
Added a
componentparameter to filter legacy weights before mapping:component="backbone": Only processes encoder/decoder weights (excludes "Head" layers)component="head": Only processes head layer weightscomponent=None: No filtering (default, for full model loading in inference)Changes
sleap_nn/legacy_models.pyfilter_legacy_weights_by_component()helper functioncomponentparameter tomap_legacy_to_pytorch_layers()andload_legacy_model_weights()sleap_nn/training/lightning_modules.pycomponent="backbone"component="head"Result
Before:
After:
Test plan
test_legacy_models.pytests pass (35 passed, 4 skipped, 3 xfailed)🤖 Generated with Claude Code