View source on GitHub
|
Module importing all agents.
Modules
behavioral_cloning module: A Behavioral Cloning agent.
categorical_dqn module: A Categorical DQN (C51) agent.
cql module: A CQL-SAC agent.
data_converter module: Agent Converter API and converters.
ddpg module: A Deep Deterministic Policy Gradient (DDPG) agent and its networks.
dqn module: A DQN (Deep Q Network) agent.
ppo module: PPO Agents.
reinforce module: A REINFORCE agent.
sac module: A Soft Actor Critic agent.
td3 module: Twin Delayed Deep Deterministic policy gradient (TD3) agent.
tf_agent module: TensorFlow RL Agent API.
Classes
class BehavioralCloningAgent: A Behavioral Cloning agent.
class CategoricalDqnAgent: A Categorical DQN Agent based on the DQN Agent.
class CqlSacAgent: A CQL-SAC Agent based on the SAC Agent.
class DdpgAgent: A DDPG Agent.
class DqnAgent: A DQN Agent.
class PPOAgent: A PPO Agent.
class PPOClipAgent: A PPO Agent implementing the clipped probability ratios.
class PPOKLPenaltyAgent: A PPO Agent implementing the KL penalty loss.
class ReinforceAgent: A REINFORCE Agent.
class SacAgent: A SAC Agent.
class TFAgent: Abstract base class for TF-based RL and Bandits agents.
class Td3Agent: A TD3 Agent.