How to save this DDPG model after the reward is saturated?

43 views Asked by At

I have built a DDPG model for my research paper. The training is going fine but I want to save the policy at the end so I don't have to train again. I have used [This DDPG link as my reference](https://keras.io/examples/rl/ddpg_pendulum/)

So the main thing I can't wrap my head around is how do I save the model as a whole. The DDPG algorithm has a critic and an actor model and a Replay Buffer. I am new to this and just want to save the model such that when I call the policy() for my action, it should come from a previously trained model.

0

There are 0 answers