Results
3
comments of
Greg
You can try to implement one of these solutions: https://stackoverflow.com/questions/33759623/tensorflow-how-to-save-restore-a-model Load a model on init() and save it every n iterations (along with logs for example).
Hey, Can you share your hyperparameters? DQN's are in general very sensitive to them and you need to finetune them carefully.
Hi, Can you share your hyperparams and loss plots? It's hard to debug your issue without them.