🗣️ Haryanvi IndicF5

Experimental Bangru Haryanvi text-to-speech.

Upload only a voice you own or have permission to clone. A clean, single-speaker 3–12 second clip works best.

Consent warning: do not impersonate people or clone a voice without the speaker's explicit authorization.

🚧 Public preview: the interface is published, but generation will be enabled only after the checkpoint passes native-speaker listening and held-out evaluation.

Model adaptation is experimental and does not represent every Haryanvi dialect.