Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix character beam decoding algorithm with vocab index map #6140

Merged
merged 1 commit into from
Mar 7, 2023

Conversation

titu1994
Copy link
Collaborator

@titu1994 titu1994 commented Mar 6, 2023

What does this PR do ?

Fixes character decoding with CTC beam search framework

Collection: [ASR]

Changelog

  • Add index vocab and vocab index maps for reverse decoding in character based models
  • Update nemo and pyctcdecode inference

Before your PR is "Ready for review"

Pre checks:

  • Make sure you read and followed Contributor guidelines
  • Did you write any new necessary tests?
  • Did you add or update any necessary documentation?
  • Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
    • Reviewer: Does the PR have correct import guards for all optional libraries?

PR Type:

  • New Feature
  • Bugfix
  • Documentation

Fixes #6127

@github-actions github-actions bot added the ASR label Mar 6, 2023
@titu1994 titu1994 requested a review from tango4j March 7, 2023 04:19
@titu1994 titu1994 merged commit dbd283c into NVIDIA:main Mar 7, 2023
@titu1994 titu1994 deleted the fix_beam_char branch March 7, 2023 19:24
titu1994 added a commit to titu1994/NeMo that referenced this pull request Mar 24, 2023
hsiehjackson pushed a commit to hsiehjackson/NeMo that referenced this pull request Jun 2, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

Successfully merging this pull request may close these issues.

eval_beamsearch_ngram.py fails (pack_hypotheses invalid data type 'str')
2 participants