#2158 File not found: C:\pyvideotrans\tmp\23676\vas-1766232874.4470487.srt

The error "File not found: C:\pyvideotrans\tmp\23676\vas-1766232874.4470487.srt" indicates that the software is looking for a temporary subtitle file that was not created. This is a common issue when the speech recognition (ASR) step fails or is interrupted.

Here are the most likely causes and solutions:

1. Speech Recognition Failure: The primary cause is that the ASR model (e.g., faster-whisper) failed to generate any subtitles. This can happen if:

   • The video/audio file contains no human speech or is mostly background noise.
   • The wrong Spoken Language was selected in the main interface (e.g., selecting Chinese for an English video).
   • The ASR model download failed or is corrupted.
   • Solution: Check the original video to confirm it contains clear speech. Ensure the correct Spoken Language is selected. If using a local model like faster-whisper, try switching to a smaller model (e.g., from large-v3 to medium) in the ASR settings.

2. Incorrect File Path/Name: The software cannot handle file paths or names containing special characters (common with videos downloaded from YouTube).

   • Solution: Rename your source video file to a short, simple name using only English letters and numbers (e.g., myvideo.mp4). Move it to a simple directory like D:\ and try processing it from there.

3. GPU/CUDA Memory Issues (OOM): If you are using GPU acceleration and the model is too large for your VRAM, the process may crash silently before creating the SRT file.

   • Solution: Go to Menu -> Tools -> Advanced Options.

     • Under Faster/OpenAI/Whisper Recognition, set CUDA Data Type to int8.
     • Set both beam_size and best_of to 1.
     • Try using the Batch Inference recognition mode in the main interface.

4. Corrupted Cache from a Previous Run: The software uses a cache. If a previous attempt failed, the cache might be causing issues.

   • Solution: On the main interface, check the box for "Clean generated files" before starting the translation process again.

Immediate Troubleshooting Steps:

1. Use the Toolbox -> Batch Speech to Subtitle function on your video file. This will test if ASR works independently. If it also fails, the issue is definitely with the ASR step.
2. Check the log files in C:\pyvideotrans\logs\ for more detailed error messages (e.g., "CUDA out of memory", "Failed to download model", etc.) and provide them if the problem persists.