def callback(info):
episode = info['episode']
params[0] = info["optimizer"].value
tracking_params[0] = info["tracking_params"]
policy_value = run_ddpg.eval_policy(callback_rngs[episode], info["optimizer"].value[0])
train_reward_per_episode.append(info["reward"])
policy_value_per_episode.append(policy_value)
elapsed_per_episode.append(info["elapsed"])
def callback(info):
episode = info['episode']
reward = info['reward']
current_actor_params = info["optimizer"].value[0]
policy_value = run_ddpg.eval_policy(callback_rngs[episode],
current_actor_params)
print(f"Episode {episode}, "
f"episode_length = {info['episode_length']}, "
f"reward = {reward}, "
f"policy_value = {policy_value}, "
f"elapsed = {info['elapsed']}")
train_reward_per_episode.append(reward)
policy_value_per_episode.append(policy_value)
episode_lengths.append(info["episode_length"])