Dataset Release Request

Hi
Its a request
Can you please release the dataset weekly or at least monthly?

1 Like

I don’t think the number of hours validated over a week or a month makes any difference to the quality of the STT models :thinking:

But once a quarter would be nice :slightly_smiling_face:
(although, again, I doubt it will make much difference)

Its depend on your views …
For example , recently , i added more than 1k voices to dataset and need my voices
Or i verified xK voices from dataset and need those verified voices
and so on
So , with current plan , i must wait for 6 or more months so i catch the result
If the release goes to every month at least , i think its okey

I know 1K voices is few, but i said for example
I can add 10K and more voices in month and need my voices for my projects

If I remember correctly from a previous post, the release process involves quite a lot of manual work. It’s not just a button press and everything gets released. To achieve faster release cycles, it would FIRST be necessary to automate a lot more things in this process.

1 Like

Oh , manual release :frowning:
I think for this type of projects , its not good :frowning:
Hope the programmers or engineers of Mozilla CommonVoice do this job and make a release schedule for at least every month :wink:

Hey @Ardin,

Thanks so much for sharing your feedback regarding dataset releases.

Given you interest in dataset releases, you might be interested in taking part in our Community Sessions next week (18th and 19th August) on the Common Voice Roadmap 2021, to learn more about the session and to register please use this link: https://ti.to/Mozilla/cv-open-roadmap-2021

1 Like