Support h5 weight loading in PyTorch training script #596
Conversation
This file contains 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
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Outdated suggestions cannot be applied.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.
Adds the ability to load initial weights from a Keras h5 model file when the --load-model argument contains the 'h5' substring.
The main changes include:
Check if the
--load-modelargument contains 'h5'. If so, callload_tf_weightsfunction to load the Keras/tf model weights into the VxmDense PyTorch model.If
--load-modelis specified but does not contain 'h5', continue to load the initial weights from a PyTorch model file.Please let me know if you have any questions or suggestions regarding this modification. Thank you for considering this pull request!