[Bug] Audio generation says "Success" in terminal but fails in UI and files are not saved

[rokonmd111]

I am encountering a critical issue while using Chatterbox-TTS. The terminal/CMD indicates that the audio generation was successful, but the UI shows a "Load Failed" error. Furthermore, no audio files are being saved in the output folder.

Steps to Reproduce:

1. Input a text longer than 5000 characters (or sometimes very short text, less than 5 characters).
2. Click on "Generate".
3. Check the Terminal: It shows "Success".
4. Check the UI: It shows "Load Failed" in the audio player area.
5. Check the output folder: No audio file is found.
6. Refresh the UI: The generated entry disappears entirely if not manually downloaded immediately (though it fails to load anyway).

Expected Behavior:

• The UI should correctly load the generated audio.
• The audio file should be permanently saved in the designated output folder automatically.
• The system should handle both long (5000+ chars) and very short texts without failing.

Actual Behavior:

• UI Issue: Shows "Load Failed" despite terminal success.
• Storage Issue: The output folder remains empty.
• Persistence Issue: Audio is lost upon refreshing the UI unless a manual download is triggered (which is currently impossible due to the load failure).

Environment:

• OS: Windows / Linux (আপনার অপারেটিং সিস্টেমের নাম লিখুন)
• GPU: RTX 5070 Ti (16GB)
• Installation Method: Local installation (Conda/venv)