Google Drive ,Saving,Loading,Resuming Features.

[kiranmaya]

Need ,these saving and loading ,resuming models states , i tried to write that , but i am confused somewhere between gpu weights ,cpu weights . or we need to create network class before loading ,or loading contains network definitions etc .

[Curt-Park]

I will give you an example of DQN's things to be considered...

1. for resuming training:

• Replay memory: self.memory
• Hyperparameters: self.batch_size, self.epsilon, self.epsilon_decay, self.max_epsilon, self.min_epsilon, self.target_update, self.gamma
• Networks: self.dqn, self.dqn_target
• Optimizer: self.optimizer

2. for saving and loading a trained model (used in test mode only):

• Hyperparameters: self.epsilon
• Networks: self.dqn

I hope you try it with the guidance above.

[Curt-Park]

Please let me know here or through another issue if you meet any other obstacles.