Zero shot voice conversion across all Indian languages, achieved by finetuning a Seed-VoiceConversion checkpoint with Indic datasets.
For instructions on local deployment and further finetuning, please refer Plachtaa/seed-vc . The finetuned checkpoints are available for download on our model page.
Note: Any reference audio will be forcefully clipped to 25s if beyond this length.
If total duration of source and reference audio exceeds 30s, source audio will be processed in chunks.