You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello, I am adapting the "stable_audio_1_0_vae.json" training config file to train the DAC autoencoder on 4 channel audio. I have mostly figured it out, but am confused about one part, the loss. I want to find spectral loss by calculating the mrstft loss of each channel individually. Im not sure if this is possible as is or if I will have to modify the source code. Thank you!
The text was updated successfully, but these errors were encountered:
Hello, I am adapting the "stable_audio_1_0_vae.json" training config file to train the DAC autoencoder on 4 channel audio. I have mostly figured it out, but am confused about one part, the loss. I want to find spectral loss by calculating the mrstft loss of each channel individually. Im not sure if this is possible as is or if I will have to modify the source code. Thank you!
The text was updated successfully, but these errors were encountered: