Fix KerasMetricCallback prediction with generate() and inference of column names #15351
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
Should now have very few differences with the PyTorch implementation
Rocketknight1
requested review from
gante,
LysandreJik,
sgugger and
merveenoyan
|
The documentation is not available anymore as the PR was closed or merged. |
sgugger
approved these changes
Jan 26, 2022
Comment on lines
+135
to
+146
| # Welp, they're not the same length. Let's do some padding | ||
| max_len = max([batch.shape[1] for batch in batches]) | ||
| num_samples = sum([batch.shape[0] for batch in batches]) | ||
| output = np.full_like( | ||
| batches[0], fill_value=padding_index, shape=[num_samples, max_len] + list(batches[0].shape[2:]) | ||
| ) | ||
| # i keeps track of which part of the concatenated array we're writing the next batch to | ||
| i = 0 | ||
| for batch in batches: | ||
| output[i : i + len(batch), : batch.shape[1]] = batch | ||
| i += len(batch) | ||
| return output |
Co-authored-by: Sylvain Gugger <[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Outdated suggestions cannot be applied.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.
A number of fixes to the Keras metric callback based on testing with the notebooks. I think I've caught all the implementation differences, and in most cases we can now pass exactly the same
compute_metricsfunction to this callback as toTrainer.I'm still in the process of overhauling the notebooks, so I might need to make a few more tweaks to this PR if I encounter any other problems!