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
Have you got observable improvements after fine-tuning with gta features? In my experiments, the quality of generated speech doesn't improve compared to the ones using the model just trained with ground truth mel-spectrogram.
The text was updated successfully, but these errors were encountered:
Have you got observable improvements after fine-tuning with gta features? In my experiments, the quality of generated speech doesn't improve compared to the ones using the model just trained with ground truth mel-spectrogram.
The text was updated successfully, but these errors were encountered: