Closed YakuzaSuske closed 1 year ago
Audio generated using Bark, (Avatar related): https://drive.google.com/file/d/1cvJuFHJHZFzdqxlRdn_oOjcoSDYT7hT8/view?usp=sharing
Note: You might hear duplicate voice lines, that's because i was testing different prompts.
Probably a duplicate of #1423
Currently you can try these: https://github.com/wsippel/bark_tts https://github.com/minemo/text-generation-webui-barktts
Probably a duplicate of #1423
Currently you can try these: https://github.com/wsippel/bark_tts https://github.com/minemo/text-generation-webui-barktts
Lol i didn't see those. My bad.
Description Bark is the new TTS and i think we would benefit from getting a TTS extension or implementation of Bark. I have played around with Bark locally on my pc and it's quite fun and sometimes realistic, almost Eleven labs level in rare cases. Tho there are some issues and fixes that would need to be implemented. The software currently by default can do a max of 15 seconds of audio, tho there is a repo that uses a code that instead cuts the prompts into pieces and basically stitches those together to make the audio longer than 15 seconds. More info on Bark below. Currently the best sounding female voice and the most natural is "en_speaker_8". The software also has voice cloning capabilities but due to misuse and etc it has been locked. Tho there is a group that managed to jailbreak it into implementing the voice cloning feature.
Additional Context Original repo: https://github.com/suno-ai/bark
The Webui that i'm using: https://github.com/Fictiverse/bark
The repo with the longer than 15 seconds limit: https://github.com/JonathanFly/bark
Jailbroken with Voice cloning: https://github.com/serp-ai/bark-with-voice-clone
Voice cloning on hugging face: https://huggingface.co/spaces/fffiloni/clone-voice-for-bark
More info on installation here: https://youtu.be/w41-MUfxIWo