# success_episodes=20
# eval_episodes=20
# iql_n_steps=100000

env,reward,success_rate,avg_return,ci_low,ci_high,across_env_mean,across_env_iqm,iqm_ci_low,iqm_ci_high

# Aggregated across envs (success rate)
reward,mean,iqm,iqm_ci_low,iqm_ci_high
