search

dunky11 / voicesmith

[WIP] VoiceSmith makes training text to speech models easy.
https://docs.voicesmith.io
Apache License 2.0
221 stars 32 forks source link

🐸Coqui <> VoiceSmith ? #6

Open erogol opened 2 years ago

erogol commented 2 years ago

Hi Tim,

I'm pinging you from 🐸 Coqui.ai and 🐸TTS.

https://github.com/coqui-ai/TTS

I just saw VoiceSmith. It is really impressive work. 🚀

I just wanted to reach out if you want to talk about a possible collaboration.

Let me know if it sounds interesting. You can reach egolge[at]coqui[dot]ai

Best :)

PS. I've tried emailing you but it didn't work.

dunky11 commented 2 years ago

I wrote you an email :)

snufas commented 1 year ago

@dunky11 Hello, amazing work, are you planning to update this repo? I would like to experiment with tts and this repo sounds very nice to try, I just can't get some things to work.

dunky11 commented 1 year ago

@snufas Hey thanks! I work full-time in text-to-speech now and developing this repository is way out of scope at the moment, unfortunately. Maybe one day I will update and finish it. I wouldn't use this project at the moment, its pretty buggy.

snufas commented 1 year ago

@dunky11 I am guessing that is Coqui.ai hence this issue... Not going to speculate any further, but nonetheless I would be really really interested to chat to you more about this. I tried the notebook you provide, it's very very good I have to say. I'm not that good at this, so I have no clue how to achieve such result with my own voice data set. No wonder why you got pinged by Coqui.ai... I am interested in text-to-speech topic particularly because I am blind myself, and I do run podcast and website for blind and visually impaired folks. Would you mind if I talk to you on email? You can email me at: vytautas@techvb.net I would really appreciate your time and expertise in this. Natural text-to-speech is a really important topic for blind and vision impaired folks like me. I am really hoping to talk to you about this topic...

Curlypla commented 1 year ago

@dunky11 I understand that you don't have the time to manage this repo with ElevenLabs anymore but could you put the dragged models online to be able to use the development version so that other people can maybe advance on their side (on google drive)? I'm mainly talking about docker v0.2.2 and the new g2p models.