Click to let us know about the errors you encounter.
3.0.0
- WhisperX, wav2vec2 word alignment and Hugging Face pyannote speaker diarization models added.
- CUDA - GPU support added for NVIDIA graphics cards. Your processes will now complete 6-8 times faster.
- Automatic subtitle generation feature added.
- Speaker recognition feature added.
- Automatic duplicate removal feature added.
- Interface additions made for subtitle and text editing.
- "Subtitles and text" category and "Notifications" category added to application settings.
- Window management features added.
- Necessary arrangements added for project directory-based operations.
- In-app informational message system added.
- Feedback system added.
2.2.2
- The transcription backend has been updated.
- The system has switched to Faster-Whisper models.
- Keyboard shortcuts have been added to the "Settings" module.
- The MurCr compatibility package has been added.
- An issue in the MurText NVDA add-on caused by changes to the WhatsApp Web interface has been fixed.
- If you are not using the classic Whisper models in other applications, follow these steps to remove them:
- Press Win + R to open the "Run" dialog.
- Paste the following line into the "Run" dialog and press Enter:
C:\Users\%username%\.cache\whisper
- Delete the packages in the opened directory.
2.1.1
- Fixed an issue where the "Default Model" could not be saved for English model packages.
- File operations have been consolidated under a single button.
- Added the ability to delete the most recent operation.
- Visual error in the update window has been fixed.
- Integrated NVDA add-on support.
- Application settings are now retained across future installations.
- Language packs have been updated.
- Added version history tracking.
2.0.2
- Expanded the user interface language support to 13 languages.
- Speech recognition now supports 100 languages.
- Application settings are now divided into two categories: "General" and "Models."
- Completed operations can now be archived if desired.
- Model download processes have been made accessible.
- Added models that recognize only English speech.
- Reworked system tray behavior.
- Added both audio and visual notifications.
- Closed the Beta 2 release cycle.