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

Update online ASR tutorial #2226

Closed
wants to merge 2 commits into from

Conversation

mthrok
Copy link
Collaborator

@mthrok mthrok commented Feb 15, 2022

https://554729-90321822-gh.circle-artifacts.com/0/docs/tutorials/online_asr_tutorial.html

  1. Add figure to explain the caching
  2. Fix the initialization of stream iterator

1. Add figure to explain the caching
2. Fix the initialization of stream iterator
Copy link
Contributor

@hwangjeff hwangjeff left a comment

Choose a reason for hiding this comment

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

just some nits — looks good

side thought: do you think it'd be helpful to be explicit about how the quality of the transcriptions is as expected partly because the model is trained on librispeech but is used to perform inference on non-librispeech data?

examples/tutorials/online_asr_tutorial.py Outdated Show resolved Hide resolved
examples/tutorials/online_asr_tutorial.py Outdated Show resolved Hide resolved
examples/tutorials/online_asr_tutorial.py Outdated Show resolved Hide resolved
examples/tutorials/online_asr_tutorial.py Outdated Show resolved Hide resolved
Co-authored-by: hwangjeff <hwangjeff@users.noreply.github.com>
@facebook-github-bot
Copy link
Contributor

@mthrok has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

@mthrok
Copy link
Collaborator Author

mthrok commented Feb 16, 2022

side thought: do you think it'd be helpful to be explicit about how the quality of the transcriptions is as expected partly because the model is trained on librispeech but is used to perform inference on non-librispeech data?

How would you phrase it?

I gave some thoughts at the very beginning, but if we mention that the sample is out-of-domain data, then the narrative is "it's not that good and the reason is the sample is out-of-domain". So I decided not to mention.

@mthrok mthrok deleted the update-online-asr-tutorial branch February 17, 2022 01:54
xiaohui-zhang pushed a commit to xiaohui-zhang/audio that referenced this pull request May 4, 2022
Summary:
https://554729-90321822-gh.circle-artifacts.com/0/docs/tutorials/online_asr_tutorial.html

1. Add figure to explain the caching
2. Fix the initialization of stream iterator

Pull Request resolved: pytorch#2226

Reviewed By: carolineechen

Differential Revision: D34265971

Pulled By: mthrok

fbshipit-source-id: 243301e74c4040f4b8cd111b363e70da60e5dae4
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants