We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
在net.py中QNetDuel、QNetTwin和QNetTwinDuel网络中都包含一个状态编码部分,但是状态编码部分不是网络的最终输出部分,为什么在状态编码后不加上激活函数呢(if_raw_out=False)?这部分是否可能会对算法性能造成影响呢,因为同样在SAC网络中也有状态编码部分,但是SAC中编码函数后却加入了激活函数。
class QNetDuel(QNetBase): # Dueling DQN def __init__(self, dims: [int], state_dim: int, action_dim: int): super().__init__(state_dim=state_dim, action_dim=action_dim) self.net_state = build_mlp(dims=[state_dim, *dims]) self.net_adv = build_mlp(dims=[dims[-1], 1]) # advantage value self.net_val = build_mlp(dims=[dims[-1], action_dim]) # Q value layer_init_with_orthogonal(self.net_adv[-1], std=0.1) layer_init_with_orthogonal(self.net_val[-1], std=0.1)
class ActorSAC(ActorBase): def __init__(self, dims: [int], state_dim: int, action_dim: int): super().__init__(state_dim=state_dim, action_dim=action_dim) self.net_s = build_mlp(dims=[state_dim, *dims], if_raw_out=False) # network of encoded state self.net_a = build_mlp(dims=[dims[-1], action_dim * 2]) # the average and log_std of action layer_init_with_orthogonal(self.net_a[-1], std=0.1)
The text was updated successfully, but these errors were encountered:
No branches or pull requests
在net.py中QNetDuel、QNetTwin和QNetTwinDuel网络中都包含一个状态编码部分,但是状态编码部分不是网络的最终输出部分,为什么在状态编码后不加上激活函数呢(if_raw_out=False)?这部分是否可能会对算法性能造成影响呢,因为同样在SAC网络中也有状态编码部分,但是SAC中编码函数后却加入了激活函数。
The text was updated successfully, but these errors were encountered: