diodiogod b13ec81872 Version 4.19.1 ...

Add CosyVoice3 <restore> tag support

Technical implementation:
- Add extract_restore_tags_only() function to filter only restore tags
- Integrate EditPostProcessor into CosyVoice processor
- Preserve CosyVoice native paralinguistic tags (<breath>, <laughter>, etc.)
- Apply ChatterBox VC restoration after TTS generation
- Block Step Audio EditX emotion/style/speed/paralinguistic tags

2026-01-28 11:43:23 -03:00

..

adapters

Add Qwen3-TTS SRT processor with character switching and language support

2026-01-28 09:11:38 -03:00

chatterbox

Fix ChatterBox model discovery to respect extra_model_paths.yaml

2025-12-29 19:35:05 -03:00

chatterbox_official_23lang

Version 4.17.0

2026-01-25 01:20:30 -03:00

cosyvoice

Version 4.16.9

2026-01-07 15:17:47 -03:00

f5_tts

Version 4.16.11

2026-01-16 00:12:51 -03:00

f5tts

Version 4.15.13

2025-12-20 11:14:31 -03:00

higgs_audio

Fix Higgs Audio numpy scope error in _prepare_inputs

2025-11-07 11:38:03 -03:00

index_tts

Version 4.16.8

2026-01-05 13:58:59 -03:00

processors

Version 4.19.1

2026-01-28 11:43:23 -03:00

qwen3_tts

Add Qwen3-TTS SRT processor with character switching and language support

2026-01-28 09:11:38 -03:00

rvc

Version 4.14.34

2025-12-02 09:54:53 -03:00

step_audio_editx

Fix Qwen3-TTS audio hash error with nested dict structure

2026-01-28 00:15:21 -03:00

vibevoice_engine

Version 4.18.0

2026-01-26 23:41:36 -03:00

video

Optimize preview video: generate directly in temp directory

2025-11-11 22:41:39 -03:00

__init__.py

Major refactoring: Reorganize project structure for better maintainability

2025-08-02 18:50:26 -03:00