#2381 ./models/iic/speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch is not registered

This error indicates that the FunASR model file is missing or corrupted. Since you are running from source on Linux, you need to download the model manually.

Solution:

1. Download the model: The required model is speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch. You can download it from the official ModelScope repository:

   # Navigate to your pyvideotrans directory
   cd /home/pietro/pyvideotrans
   # Create the models directory if it doesn't exist
   mkdir -p ./models/iic
   # Download the model using git (recommended for large files)
   git lfs install
   git clone https://www.modelscope.cn/iic/speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch.git ./models/iic/speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch

   If you don't have git-lfs installed, you can download the .tar.gz file from the ModelScope page and extract it to the ./models/iic/ directory.

2. Verify the path: Ensure the final model folder structure looks like this:

   /home/pietro/pyvideotrans/models/iic/speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch/
       ├── am.mvn
       ├── configuration.json
       ├── ... (other model files)

3. Restart the application: After the model is downloaded, restart pyVideoTrans. The error should be resolved.

Alternative Quick Fix:
If you don't specifically need the superior Chinese recognition of FunASR, you can simply switch your Speech Recognition method to faster-whisper(local) in the main interface, which will use a different, automatically-downloaded model.

Please refer to the documentation for more details:

• 阿里FunASR中文识别