Closed Pythonpa closed 5 months ago
Hey, thank you for pointing this out, this is not stuck, in gradio UI, you can see the step count increasing. Progress bar will be implemented soon in process image tab too, so that this doesn't happen again. Also, to get better results I would recommend using weight lower than 0.5 when using codeformer and weight >= 0.5 when using GFPGAN, I will soon be updating the wav2lip model, these artifacts emerge because wav2lip is trained in low quality faces.
Hey, thank you for pointing this out, this is not stuck, in gradio UI, you can see the step count increasing. Progress bar will be implemented soon in process image tab too, so that this doesn't happen again. Also, to get better results I would recommend using weight lower than 0.5 when using codeformer and weight >= 0.5 when using GFPGAN, I will soon be updating the wav2lip model, these artifacts emerge because wav2lip is trained in low quality faces.
OK,will have a try again.........
I have updated it to now show "Processing..." and also print "Done" when it's done.
Hi,thanks for your excellent work.When I uploaded my video and sound,I got a weired result,It just stucking,no other info-out:
"" Getting face landmarks... W0000 00:00:1705849449.547341 19148 face_landmarker_graph.cc:174] Sets FaceBlendshapesGraph acceleration to xnnpack by default. Using cuda for inference. Generating audio spectrogram... Video is longer than audio. Truncating video... Length of mel chunks: 855 Load GFPGAN checkpoint from: D:\PythonProject\Lip_Wise\weights\gfpgan\GFPGANv1.4.pth Loading wav2lip checkpoint from: D:\PythonProject\Lip_Wise\weights\wav2lip\wav2lip.pth ""
How to fix the problem?