Mirrors/bark-with-voice-clone
1
0
Fork 0
mirror of https://github.com/serp-ai/bark-with-voice-clone.git synced 2025-12-15 03:07:58 +01:00
Code Issues Packages Projects Releases Wiki Activity

Update README.md

Browse Source
This commit is contained in:
Francis LaBounty
2023-05-25 14:38:35 -06:00
committed by GitHub
parent 2ccbeb5635
commit 0b16a49fe2
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
README.md
View File

@@ -2,9 +2,9 @@
If you want to clone a voice just follow the `clone_voice.ipynb` notebook. If you want to generate audio from text, follow the `generate.ipynb` notebook.
To create a voice clone sample, you need an audio/text pair of less than 7 seconds. (limited testing shows better results with shorter samples (2-4 seconds))
To create a voice clone sample, you need an audio/text pair of less than 13 seconds.
Haven't experimented with what kind of audio/text pairs work best, but this will be updated as we find out more.
You will get the best results by making generations with your cloned voice until you find one that is really close to the source. Then use that as the new history prompt (comes from the model so should theoretically be more consistent)
- [BARK text to speech @ SERP AI](https://serp.ai/tools/bark-text-to-speech-ai-voice-clone-app/)