ef18893fb50c97c2c0d9d7b8b951455683f89635,rllib/agents/ppo/appo_tf_policy.py,,stats,#Any#Any#,342

Before Change

    if policy.config["vtrace"]:
        is_stat_mean, is_stat_var = tf.nn.moments(policy.loss.is_ratio, [0, 1])
        stats_dict.update({"mean_IS": is_stat_mean})
        stats_dict.update({"var_IS": is_stat_var})

    if policy.config["use_kl_loss"]:
        stats_dict.update({"kl": policy.loss.mean_kl})
        stats_dict.update({"KL_Coeff": policy.kl_coeff})

    return stats_dict

After Change

        stats_dict["var_IS"] = is_stat_var

    if policy.config["use_kl_loss"]:
        stats_dict["kl"] = policy._mean_kl
        stats_dict["KL_Coeff"] = policy.kl_coeff

    return stats_dict

Instances