diodiogod f35395a072 Version 4.8.30 ...

Fix VibeVoice cache invalidation and DynamicCache compatibility

Technical details:
- Add inference_steps parameter to VibeVoice cache key generator
- Disable deprecated transformers compatibility patches causing DynamicCache errors
- Add targeted DynamicCache key_cache/value_cache properties for VibeVoice compatibility
- Fix cache not invalidating when changing diffusion inference steps

2025-09-09 14:35:23 -03:00

..

audio

Version 4.8.30

2025-09-09 14:35:23 -03:00

compatibility

Version 4.8.30

2025-09-09 14:35:23 -03:00

downloads

Add ChatterBox Official 23-Lang multilingual TTS engine

2025-09-04 19:11:49 -03:00

models

Version 4.8.27

2025-09-09 00:40:06 -03:00

streaming

Fix audio segments/timings mismatch in SRT streaming

2025-08-25 22:48:15 -03:00

system

Version 4.5.13: Clean up excessive startup logging messages

2025-08-26 09:47:39 -03:00

text

Add Italian to F5-TTS phonemization exceptions

2025-09-06 14:13:34 -03:00

timing

Fix tensor dimension issues in concatenation assembly

2025-09-05 20:33:06 -03:00

voice

Version 4.8.28

2025-09-09 11:03:05 -03:00

__init__.py

Major refactoring: Reorganize project structure for better maintainability

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

phoneme_matcher.py

Fix word prediction blank elimination and improve repetitive pattern variety

2025-08-15 12:12:42 -03:00

robust_import.py

Version 4.6.8

2025-08-30 00:28:43 -03:00