Question

class TD3(model, gamma=None, tau=None, actor_lr=None, critic_lr=None, policy_noise=0.2, noise_clip=0.5, policy_freq=2)[源代码]¶

基类：Algorithm

__init__(model, gamma=None, tau=None, actor_lr=None, critic_lr=None, policy_noise=0.2, noise_clip=0.5, policy_freq=2)[源代码]¶

TD3 algorithm

参数:

• model (parl.Model) – forward network of actor and critic.

• gamma (float) – discounted factor for reward computation

• tau (float) – decay coefficient when updating the weights of self.target_model with self.model

• actor_lr (float) – learning rate of the actor model

• critic_lr (float) – learning rate of the critic model

• policy_noise (float) – noise added to target policy during critic update

• noise_clip (float) – range to clip target policy noise

• policy_freq (int) – frequency of delayed policy updates

learn(obs, action, reward, next_obs, terminal)[源代码]¶

Define the loss function and create an optimizer to minize the loss.

predict(obs)[源代码]¶

Refine the predicting process, e.g,. use the policy model to predict actions.