[ci-daily] Fix pipeline tests #21257
Merged
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
|
The documentation is not available anymore as the PR was closed or merged. |
ArthurZucker
commented
Jan 23, 2023
Comment on lines
-638
to
+645
| word_offsets = [] | ||
| offsets = [] | ||
| for word, (start_offset, end_offset) in chunk_offset: | ||
| word_offsets.append({"word": word, "start_offset": start_offset, "end_offset": end_offset}) | ||
| offsets.append({"word": word, "start_offset": start_offset, "end_offset": end_offset}) |
Collaborator
Author
There was a problem hiding this comment.
normalized the name of this variable
| yield {"is_last": True, **processed, **extra} | ||
|
|
||
| def _forward(self, model_inputs, generate_kwargs=None): | ||
| def _forward(self, model_inputs, return_timestamps=False, generate_kwargs=None): |
Collaborator
Author
There was a problem hiding this comment.
Adding this argument prevent the .pop from removing it for other processes.
Comment on lines
-104
to
+113
| consecutive = np.where(timestamp_tokens[:-1] & timestamp_tokens[1:])[0] + 1 | ||
| last_timestamp = np.where(timestamp_tokens)[0][-1] | ||
| consecutive = np.append(consecutive, last_timestamp) if last_timestamp not in consecutive else consecutive | ||
| if seq_idx != 0: | ||
| if seq_idx != 0 and sum(timestamp_tokens) > 0: | ||
| consecutive = np.where(timestamp_tokens[:-1] & timestamp_tokens[1:])[0] + 1 | ||
| last_timestamp = np.where(timestamp_tokens)[0][-1] | ||
| consecutive = np.append(consecutive, last_timestamp) if last_timestamp not in consecutive else consecutive |
Collaborator
Author
There was a problem hiding this comment.
This just makes sure that if the model output no timestamps, we just don't throw an error
Comment on lines
+74
to
+79
| if "input_features" in processed: | ||
| processed_len = processed["input_features"].shape[-1] | ||
| elif "input_values" in processed: | ||
| processed_len = processed["input_values"].shape[-1] | ||
| if processed_len != chunk.shape[-1] and rescale: | ||
| ratio = processed_len / chunk_len |
Collaborator
Author
There was a problem hiding this comment.
This was missing! Fixes the LM tests
Narsil
approved these changes
Jan 23, 2023
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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.
What does this PR do?
Should fix the
automatic_speech_recognition_pipelinetests.Also using
streamingdataset to speed up tests. Think it is a good idea if we are only using 1 data.