This is a web, Android and iOS app for collecting speech donations for the Common Voice project.
By participating in this project, you're agreeing to uphold the Mozilla Community Participation Guidelines. If you need to report a problem, please see our REPORTING.md guide.
From writing and reading sentences to enhancing our front-end architecture, there are many ways to get involved with Common Voice. For more information, check out CONTRIBUTING.md.
For general discussion (feedback, ideas, random musings), head to our Discourse Category.
For technical problems or suggestions, please use the GitHub issue tracker.
Or come chat with us on Slack: Invite Link
- yarn install
- npm install
- place a config.json under server/ with db parameters and s3 parameters as described in docs/HOWTO_S3.md
you will need to configure a root password for mysql as well, so migrations can be processed
- npm run build
- npm run start:prod
- you are now up and running on localhost:9000
If you want to change transcript data you can find them under server data.
Data in your S3 bucket will be stored under a uuid per user(folder) and track (file.mp3) with a belonging .txt file which stores the transcript.
To import your voice-corpus with mozilla-deepspeech you'll need to tar.gz your bucket. After this you can use: https://github.com/mozilla/DeepSpeech/blob/master/bin/import_cv.py to import it.