Deep Reinforcement Learning Hands-On

Training code

We have two very similar training modules in this example: one for the feed-forward model and one for 1D convolutions. For both of them, there is nothing new added to our examples from Chapter 7, DQN Extensions:

• They’re using epsilon-greedy action selection to perform exploration. The epsilon linearly decays over the first 1M steps from 1.0 to 0.1.
• A simple experience replay buffer of size 100k is being used, which is initially populated with 10k transitions.
• For every 1000 steps, we calculate the mean value for the fixed set of states to check the dynamics of the Q-values during the training.
• For every 100k steps, we perform validation: 100 episodes are played on the training data and on previously unseen quotes. Characteristics of ...