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

[bugfix] avoid the random shuffle of phoneme and tone tokens. #6855

Merged
merged 4 commits into from
Jun 17, 2023

Conversation

XuesongYang
Copy link
Collaborator

@XuesongYang XuesongYang commented Jun 12, 2023

What does this PR do ?

The order of outputs of dict.values() is usually shuffled so that the tokenizer and the model will be confused to match a correct sound for a specific integer index. You can tell the diff in between as below,

CTC Training Loss Curves
The upper and bottom curves are referring to the current implementation and bugfix, respectively.
W B Chart 6_12_2023, 12_08_47 PM

Phoneme Token List

  • current random order: ['#Q', '#VE', '#IE', '#X', '#Y', '#ZH', '#S', '#EI', '#ANG', '#V', '#O', '#H', '#D', '#F', '#T', '#G', '#OU', '#UI', '#UN', '#B', '#IU', '#I', '#C', '#AO', '#IN', '#L', '#^', '#M', '#ING', '#VN', '#Z', '#SH', '#EN', '#J', '#P', '#K', '#CH', '#AI', '#A', '#ENG', '#R', '#W', '#U', '#E', '#AN', '#N', '#ER', '#ONG']
  • bugfix: ['#A', '#AI', '#AN', '#ANG', '#AO', '#B', '#C', '#CH', '#D', '#E', '#EI', '#EN', '#ENG', '#ER', '#F', '#G', '#H', '#I', '#IE', '#IN', '#ING', '#IU', '#J', '#K', '#L', '#M', '#N', '#O', '#ONG', '#OU', '#P', '#Q', '#R', '#S', '#SH', '#T', '#U', '#UI', '#UN', '#V', '#VE', '#VN', '#W', '#X', '#Y', '#Z', '#ZH', '#^']

Audio Examples

Collection: [Note which collection this PR will affect]

Changelog

  • Add specific line by line info of high level changes in this PR.

Usage

  • You can potentially add a usage example below
# Add a code snippet demonstrating how to use this 

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

If you haven't finished some of the above items you can still open "Draft" PR.

Who can review?

Anyone in the community is free to review the PR once the checks have passed.
Contributor guidelines contains specific people who can review PRs to various areas.

Additional Information

  • Related to # (issue)

Copy link
Collaborator

@redoctopus redoctopus left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for catching & fixing this! LGTM.

@XuesongYang XuesongYang merged commit 6f2035b into NVIDIA:main Jun 17, 2023
11 of 13 checks passed
@XuesongYang XuesongYang deleted the bugfix-dict-order branch June 17, 2023 23:39
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.

None yet

3 participants