If there is a direct configuration available to aggregate all episodes-IDs by iteration and subsequently log them at the end of each iteration in PPOTraining? or do I have to write a CustomPPOTrainer(PPOTrainer) by inheriting PPOTrainer. ?
Related topics
Topic | Replies | Views | Activity | |
---|---|---|---|---|
Can ray allow access to individual episodes? | 5 | 445 | September 22, 2021 | |
How to tell RLLIB trainer (Not Tune) to run that many number of episodes | 7 | 1159 | June 9, 2023 | |
[Rllib] Store actions during training with PPOTrainer to get statistics about action-distribution over episodes | 1 | 475 | October 21, 2022 | |
[RLlib, Tune, PPO] episode_reward_mean based on new episodes for each iteration | 1 | 26 | November 25, 2024 | |
Accessing custom metrics for episodes | 3 | 797 | March 19, 2024 |