`exarl.agents.agent_vault.ddpg_vtrace`

Module Contents

Classes

Agent base class: Inherits from abstract base class for mandating

Functions

update_target(target_weights, weights, tau)

Attributes

exarl.agents.agent_vault.ddpg_vtrace.logger

exarl.agents.agent_vault.ddpg_vtrace.update_target(target_weights, weights, tau)

class exarl.agents.agent_vault.ddpg_vtrace.DDPG_Vtrace(env, is_learner)

Bases: exarl.ExaAgent

Agent base class: Inherits from abstract base class for mandating functionality (pure virtual functions).

is_learner :bool

remember(self, state, action, reward, next_state, done)

update_grad(self, state_batch, action_batch, reward_batch, next_state_batch)

get_actor(self)

get_critic(self)

generate_data(self)

train(self, batch): train the agent

target_train(self)

action(self, state): next action based on current state

get_weights(self): get target model weights

set_weights(self, weights): set target model weights

set_learner(self)

update(self)

load(self, filename): load weights

save(self, filename): save weights

monitor(self)

set_agent(self)

epsilon_adj(self)