Mombin

you can run fix this argument python test_all.py --test_root ./0572_0019_0003/audio --test_type audio --test_audio_video_length 99 --test_resume_path CHECKPOINT_PATH -----> python test_all.py --test_root ./0572_0019_0003/audio --test_type audio --test_audio_video_length 98 --test_resume_path CHECKPOINT_PATH

modify your code in embedding_utils.py origin: import Options config = Options.Config() revise: from Options import Config config = Config().parse() and if you find same error in another code, you can...

@hunghn-dmobin Insert the following keywords into the introduction of your code. torch.cuda.set_device(device) then it is working fine.