tugstugi/pytorch-dc-tts

Transfer learning from already trained models

subramaniannk opened this issue · 1 comments

I am having some difficulties in loading already trained models. Could you be more specific how I can load already trained models? What are the changes that are to be made in the code?

If you are training English, you can load the LJSpeech model directly and fine tune. If not, it is little complicated you have to remove the embedding layer from the pretrained LJSpeech text2mel.