tf_agents.agents.reinforce.reinforce_agent.ReinforceAgentLossInfo

Stay organized with collections Save and categorize content based on your preferences.

ReinforceAgentLossInfo is stored in the extras field of the LossInfo.

tf_agents.agents.reinforce.reinforce_agent.ReinforceAgentLossInfo(
    policy_gradient_loss,
    policy_network_regularization_loss,
    entropy_regularization_loss,
    value_estimation_loss,
    value_network_regularization_loss
)

All losses, except for policy_network_regularization_loss have a validity mask applied to ensure no loss or error is calculated for episode boundaries.

policy_gradient_loss: The weighted policy_gradient loss. policy_network_regularization_loss: The regularization loss terms from the policy network used to generate the policy_gradient_loss. entropy_regularization_loss: The entropy regularization loss. value_estimation_loss: If value estimation network is being used, the loss associated with that network.

policy_gradient_loss A namedtuple alias for field number 0
policy_network_regularization_loss A namedtuple alias for field number 1
entropy_regularization_loss A namedtuple alias for field number 2
value_estimation_loss A namedtuple alias for field number 3
value_network_regularization_loss A namedtuple alias for field number 4