Beyond the Boundaries of Proximal Policy Optimization
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913569782628352 |
|---|---|
| author | Tan, Charlie B. Toledo, Edan Ellis, Benjamin Foerster, Jakob N. Huszár, Ferenc |
| author_facet | Tan, Charlie B. Toledo, Edan Ellis, Benjamin Foerster, Jakob N. Huszár, Ferenc |
| contents | Proximal policy optimization (PPO) is a widely-used algorithm for on-policy reinforcement learning. This work offers an alternative perspective of PPO, in which it is decomposed into the inner-loop estimation of update vectors, and the outer-loop application of updates using gradient ascent with unity learning rate. Using this insight we propose outer proximal policy optimization (outer-PPO); a framework wherein these update vectors are applied using an arbitrary gradient-based optimizer. The decoupling of update estimation and update application enabled by outer-PPO highlights several implicit design choices in PPO that we challenge through empirical investigation. In particular we consider non-unity learning rates and momentum applied to the outer loop, and a momentum-bias applied to the inner estimation loop. Methods are evaluated against an aggressively tuned PPO baseline on Brax, Jumanji and MinAtar environments; non-unity learning rates and momentum both achieve statistically significant improvement on Brax and Jumanji, given the same hyperparameter tuning budget. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2411_00666 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Beyond the Boundaries of Proximal Policy Optimization Tan, Charlie B. Toledo, Edan Ellis, Benjamin Foerster, Jakob N. Huszár, Ferenc Machine Learning Artificial Intelligence Proximal policy optimization (PPO) is a widely-used algorithm for on-policy reinforcement learning. This work offers an alternative perspective of PPO, in which it is decomposed into the inner-loop estimation of update vectors, and the outer-loop application of updates using gradient ascent with unity learning rate. Using this insight we propose outer proximal policy optimization (outer-PPO); a framework wherein these update vectors are applied using an arbitrary gradient-based optimizer. The decoupling of update estimation and update application enabled by outer-PPO highlights several implicit design choices in PPO that we challenge through empirical investigation. In particular we consider non-unity learning rates and momentum applied to the outer loop, and a momentum-bias applied to the inner estimation loop. Methods are evaluated against an aggressively tuned PPO baseline on Brax, Jumanji and MinAtar environments; non-unity learning rates and momentum both achieve statistically significant improvement on Brax and Jumanji, given the same hyperparameter tuning budget. |
| title | Beyond the Boundaries of Proximal Policy Optimization |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2411.00666 |