(self, state_size, action_size)
| 246 | |
| 247 | class DQNAgent(object): |
| 248 | def __init__(self, state_size, action_size): |
| 249 | self.state_size = state_size |
| 250 | self.action_size = action_size |
| 251 | self.memory = ReplayBuffer(state_size, action_size, size=500) |
| 252 | self.gamma = 0.95 # discount rate |
| 253 | self.epsilon = 1.0 # exploration rate |
| 254 | self.epsilon_min = 0.01 |
| 255 | self.epsilon_decay = 0.995 |
| 256 | self.model = mlp(state_size, action_size) |
| 257 | |
| 258 | |
| 259 | def update_replay_memory(self, state, action, reward, next_state, done): |
nothing calls this directly
no test coverage detected