Bluebomber182's picture
Update README.md
429dfc2 verified
|
raw
history blame
284 Bytes
metadata
license: cc-by-nc-4.0

This was trained on the Emilia Dataset and trimed down Emilia-YODAS and AniSpeech datasets that pass the 3.6 mos score threshold. This has f0 condition set to true so you can app_svc.py on it. Note it has an inference problem with any of the checkpoints.