Support for returning the hidden vector in the Wav2Vec2 and Wav2Vec2Bert models#1867
Merged
Conversation
Contributor
Author
added 2 commits
March 14, 2025 21:57
Collaborator
|
Hello, Can you rebase on master to see ìf the pipeline is fixed? |
Contributor
Author
|
@minhthuc2502 all passed indeed. Thank you so much! Would you please merge it? |
Collaborator
|
Look good to me. Thank you! |
Lanurence666
pushed a commit
to Lanurence666/CTranslate2_backport_py38
that referenced
this pull request
May 23, 2026
…ert models (OpenNMT#1867) * adding the hidden vector return feature * format update * format update * dummy * dummy --------- Co-authored-by: hkwon <[email protected]>
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This PR adds support for returning the hidden vector in the Wav2Vec2 and Wav2Vec2Bert models. By integrating these models into the CTranslate2 framework, this enhancement enables more efficient application in speech encoder tasks, particularly for low-resource systems. To activate this feature, simply add
"return_hidden": true,to the config.json. This modification detaches the lm_head during the model conversion and returns the last hidden vector of the model.