HH  rA brain.Event:2R. ,tensorboard.summary.writer.event_file_writerM"x= 2rA*  objective/kl,\'F rA*  objective/kl_coefL>!'F 9rA*  objective/entropy̿Bdfo/m]P PrA*" ppo/mean_non_score_reward"uE%6 frA*  ppo/mean_scoresU}=#jW$B+M yrA*  ppo/std_scoresz>&G-F5(pJ  rA*  ppo/policy/entropyW@t-)7_ ? rA*  ppo/policy/approxkl >oG38)7_  rA*  ppo/policy/policyklŌ= )7_ rA*  ppo/policy/clipfrac>6"0_ W rA*# ! ppo/policy/advantages_mean̯Mi&sO rA*  ppo/returns/meanB?%6 rA*  ppo/returns/varԙ>ÈƂ#wC LrA*  ppo/val/vpred:@EM#wC rA*  ppo/val/errorϸ@&sO rA*  ppo/val/clipfracf?l"x= rA*  ppo/val/meanvS@Q3!{ rA*  ppo/val/var݋8@D(+K ZrA*  ppo/val/var_explainedwFaB'F rA*  ppo/learning_rate l7!k+K rA*  time/ppo/forward_passp;?>J3.W 9rA*!  time/ppo/compute_rewardsz;8s1 frA*$ " time/ppo/compute_advantagesLQ;&^,E rA*  time/ppo/optimize_step<ρ?= {)7_ rA*  time/ppo/calc_statsЅ0>'8z$B+M rA*  time/ppo/totalz?&%6 rA*  env/reward_meanU}=K$B+M rA*  env/reward_stdz>a