How to use train model on Google Colab with "CREPE" mode.

[jonnytri53]

This is an original code from Google Colab. How can I modify the code similarly like it in the svcg-gui on Windows?. Thank you.

#@title Use trained model
#@markdown **Put your .wav file in `so-vits-svc-fork/audio` directory**
from IPython.display import Audio, display
!svc infer drive/MyDrive/so-vits-svc-fork/audio/{NAME}.wav -m drive/MyDrive/so-vits-svc-fork/logs/44k/ -c drive/MyDrive/so-vits-svc-fork/logs/44k/config.json
display(Audio(f"drive/MyDrive/so-vits-svc-fork/audio/{NAME}.out.wav", autoplay=True))

[34j]

Read the docs

[jonnytri53]

Hi @34j
I have attempted to go through the readme document, but I couldn't find any information about incorporating the Crepe mode in Google Colab. As I'm relatively new to this, I apologize for any inconvenience. Could you please take a moment to review the original code and guide me on how to make the necessary modifications? I would appreciate it so much.
This is my modification code, but it wasn't working.

#@title Use trained model
#@markdown Put your .wav file in so-vits-svc-fork/audio directory
from IPython.display import Audio, display
!svc infer -fm crepe drive/MyDrive/so-vits-svc-fork/audio/jonnytri.wav -m drive/MyDrive/so-vits-svc-fork/logs/44k/ -c drive/MyDrive/so-vits-svc-fork/logs/44k/config.json
display(Audio(f"drive/MyDrive/so-vits-svc-fork/audio/jonnytri.out.wav", autoplay=True))
Sorry for my inconvenience.

[34j]

That should be correct...but I need more information.