v0.7.1 - Important Bugfixes

Bugfix: Clone Voice missing tokenizer selection
Bugfix: Installer not cloning/pulling source
Installer always does a hard reset to ignore local changes

v0.7.0 - Huge Improvements and voice cloning/swapping

• Implemented new voice cloning method, made possible by Mylo
• Added Swap Voice Tab, to swap voices with selected speakers in input audio file, also made possible by the use of Hubert
• Added Training Preparation & Training Tabs to train quantifiers for your own native language (WIP)
• Trained my own german quantifier/tokenizer which will be downloaded/used when you're selecting german as base language for cloning
• Moved extra dependencies into separate requirements.txt allowing a simpler, cleaner install
• Rewrote parts of the 1-click installer due to that reason
• Refactored lots of stuff

v0.4.7: - BUGFIX: 1-click installer, several changes

• BUGFIX: 1-click installer, several changes
• BUGFIX: New commandline args weren't used
• Commandline-Args possible in windows installer

and all improvements from v0.4.6

v0.4.5

• BUGFIX: Blank Voice Lines when splitting for SSML
• Added Settings Tab
• Added Theming Support
• Added initial Generation Seeds for reproducing results
• ID3 Tag with generation infos written to WAV
• Streamlined prompt settings, removed checkboxes for history prompts and added slider for min_eos_p

v0.4.0 - First Release with Windows Installer

- BUGFIX: Joining prompt names in subfolders

- Release with Windows Installer
- Installer .bat includes automatic updater
- Trying to detect voice changes, adding short pause breaks inbetween