Fine tuning?
#18
by
potatocharlie
- opened
Hey there, I know this might be a stupid question, but how to fine-tune this model? Or is it even possible? Thanks in advance!
Hey
@potatocharlie
- you can follow the steps for training the FastSpeech2 LJSpeech checkpoint on the fairseq repo: https://github.com/facebookresearch/fairseq/blob/main/examples/speech_synthesis/docs/ljspeech_example.md#ljspeech
Note that these instructions cover how to pre-train the model on the LJSpeech dataset. If you want to fine-tune it, you'll need to adapt the provided examples for your dataset.