classic DQN 코드 구현2025.03.31 - [코딩] - DQN 코드 구현 DQN 코드 구현Replay Bufferclass ReplayBuffer: def __init__(self, buffer_size, state_dim, action_dim, device="cpu"): self.buffer_size = buffer_size self.state_dim = state_dim self.action_dim = action_dim self.device = device self.observations = np.zeros((buffer_size, *soccident.tistory.com  class QNetwork(nn.Module): def __init__(self, env): su..

classic DQN 코드 구현2025.03.31 - [코딩] - DQN 코드 구현 DQN 코드 구현Replay Bufferclass ReplayBuffer: def __init__(self, buffer_size, state_dim, action_dim, device="cpu"): self.buffer_size = buffer_size self.state_dim = state_dim self.action_dim = action_dim self.device = device self.observations = np.zeros((buffer_size, *soccident.tistory.com  Double DQN의 경우 가장 코드 구현이 쉬웠다.obs, actions, next_obs, rewards, do..

Classic DQN2025.03.31 - [코딩] - DQN 코드 구현 DQN 코드 구현Replay Bufferclass ReplayBuffer: def __init__(self, buffer_size, state_dim, action_dim, device="cpu"): self.buffer_size = buffer_size self.state_dim = state_dim self.action_dim = action_dim self.device = device self.observations = np.zeros((buffer_size, *soccident.tistory.com Prioritized Experience Replayclass PrioritizedReplayBuffer: def __in..

Replay Bufferclass ReplayBuffer: def __init__(self, buffer_size, state_dim, action_dim, device="cpu"): self.buffer_size = buffer_size self.state_dim = state_dim self.action_dim = action_dim self.device = device self.observations = np.zeros((buffer_size, *state_dim), dtype=np.uint8) self.next_observations = np.zeros((buffer_size, *state_dim), dtype=np...