sinergym.utils.evaluation

Custom policy evaluations for Evaluation Callbacks.

Functions

evaluate_policy(model, env[, ...])

Runs policy for n_eval_episodes episodes and returns average reward.