Fine tuning a preexisting model

fatchord / WaveRNN

WaveRNN Vocoder + TTS

https://fatchord.github.io/model_outputs/

MIT License

2.14k stars 698 forks source link

Fine tuning a preexisting model #166

Open Jasondhillon opened 4 years ago

Jasondhillon commented 4 years ago

Hi there, How would one go about fine tuning the preexisting models to a dataset of my own?

lenbrocki commented 4 years ago

I am also interested in fine-tuning the model. More specifically, since there is no parameter for this, how can I freeze the encoder to only finetune the decoder with a small dataset?