Estimator loading from file

[MeFistoPH3L]

First of all, let me thank the community for the work done. You are incredible!

My question is: can I save the DeepAR model after training to a file, and then when I run the script again, load this model and continue training it?

From what I found myself is this:

#saving predictor
model.serialize(Path(directory))

#load predictor after
model = Predictor.deserialize(Path(directory))

But unfortunately the predictor cannot be trained, so this option is not suitable