Can someone post python code to save a model once trained and then load it for actual prediction?
1 Like
Hey @Hannan , could you take a look at the scripts in here:
There are also newer ONNX examples for both tf and torch, if you’d like to use ONNX as your export format.
I looked at the examples. This is quite complicated. I am using stable baselines (3) now and it takes only one line of code to save/load model with all dependencies. Is there a shorter way in rllib?
1 Like