The training code is now live at https://github.com/kyutai-labs/pocket-tts/tree/main/training !
EDIT: The 3.0.0 wheel is broken due to a bad build, use 3.0.1. The 3.0.0 wheel has been yanked anyway.
What's Changed
- Fix pocket-tts generate docs by @gabrieldemarmiesse in #182
- Add new macOS app links to README by @slaughters85j in #193
- Add 'seshat-tts' project to README by @Iheuzio in #197
- Update README.md by @joshwhiton in #200
- Default the English model's temperature to 0.3 by @manukyutai in #223
- Fix audio and state edge cases by @ShiroKSH in #210
- Fix docker compose command duplicating entrypoint args by @grigio in #233
- docs: correct GPU speedup claim + add CUDA usage notes by @moduvoice in #213
- docs: quote pip extras install examples by @nyxst4ck in #208
- Rename modules: mimi_transformer -> transformer, transformer -> attention by @manu-lm in #237
- Mimi: decode_from_latent owns quantizer + transposes; encode_to_latent returns [B, T, C] by @manu-lm in #239
- Add the training code by @manukyutai in #244
- Add the 24l variant of the english model by @gabrieldemarmiesse in #245
New Contributors
- @slaughters85j made their first contribution in #193
- @Iheuzio made their first contribution in #197
- @manukyutai made their first contribution in #223
- @ShiroKSH made their first contribution in #210
- @grigio made their first contribution in #233
- @moduvoice made their first contribution in #213
- @nyxst4ck made their first contribution in #208
- @manu-lm made their first contribution in #237
Full Changelog: v2.1.0...v3.0.0