3 ms·
My language has under 1000 hours of training data, so apparently is not well supported. How can I help add more training data? I actually have tens of hours of
by dotancohen 2y ago
My language has under 1000 hours of training data, so apparently is not well supported. How can I help add more training data? I actually have tens of hours of transcriptions of my own voice in my language, because I take many voice notes spanning back almost two decades. Most of it is very personal, but I could probably sort away a good portion for this and other projects.
https://keyboard.futo.org/whisper-training-data-breakdown https://keyboard.futo.org/whisper-training-data-breakdown
- richardboegli 2y agoHi dotancohen, More info of how to help here. https://github.com/futo-org/voice-input-models https://github.com/futo-org/voice-input-models
- dotancohen 2y agoThank you!
- yorwba 2y agoIf you contribute to Mozilla Commmon Voice, it'll probably make its way into their training data eventually. https://commonvoice.mozilla.org/en/languages https://commonvoice.mozilla.org/en/languages If you want to use your existing recordings, uploading them to YouTube or archive.org with a Creative Commons license might also work.
- dotancohen 2y agoGreat, thanks, I'll spend some time with that.