Fine-tuning Magneta GrooVAE

[Yair Hashachar]

Hello everyone,

I'm reaching out for some assistance regarding fine-tuning a pre-trained checkpoint with my own data. I have successfully trained a model from scratch, but I'm unsure about the best approach for fine-tuning an existing checkpoint. Should I simply load the checkpoint as a starting point, or is there more to it?

I would greatly appreciate any guidance you can provide.

Thank you, 

Yair