#3976 Save audio in G:/pyvideo/f5-tts

The issue is that the F5-TTS model files are not being found in the expected directory structure.

To resolve this, you need to manually download and place the F5-TTS model files in the correct folder within your pyVideoTrans installation.

1. Create the correct folder path: Navigate to your pyVideoTrans installation directory (e.g., G:/pyvideo/). Inside it, go to the models folder. Create a new folder named exactly: models--f5-tts--f5-tts-v2
   The full path should be: G:/pyvideo/models/models--f5-tts--f5-tts-v2
2. Download the model files: Visit the F5-TTS model page on HuggingFace: https://huggingface.co/f5-tts/f5-tts-v2/tree/main
3. Place the files: Download all the .json, .bin, .txt, and other files from that page. Copy and paste these downloaded files directly into the models--f5-tts--f5-tts-v2 folder you created in step 1.
4. Restart and configure: After placing the files, restart pyVideoTrans (sp.exe). Ensure you have the F5-TTS API service running (typically at http://127.0.0.1:7860). In the software's main menu, go to Settings -> TTS Settings -> F5-TTS and enter the correct API address.

Please refer to the relevant documentation:

• F5-TTS api调用常见错误
• 使用F5-TTS配音