Open enochkan opened 3 years ago
There are no training epochs as the data the model is training on is constantly changing and dependent on actions the model takes.
The model is setup to save during training when the Test Agent process completes a game and achieves a new high score or matches high score it has achieved on game.
Just want to clarify that there is only one saved model per environment and it will be overwritten each training epoch, right? For example, MsPacman will only have one saved model
trained_models/MsPacman-v0.dat