Reinforcement learning class (RofuncRL)#
1. Online reinforcement learning#
- Rofunc RL
Overview of the RofuncRL subpackage
- RofuncRL A2C (Advantage Actor-Critic)
A2C implementation and explanation of tricks in RofuncRL
- RofuncRL PPO (Proximal Policy Optimization)
PPO implementation and explanation of tricks in RofuncRL
- RofuncRL TD3 (Twin Delayed Deep Deterministic Policy Gradient)
TD3 implementation and explanation of tricks in RofuncRL
- RofuncRL SAC (Soft Actor-Critic)
SAC implementation and explanation of tricks in RofuncRL
2. Offline reinforcement learning#
- RofuncRL CQL (Conservative Q-Learning)
CQL implementation and explanation of tricks in RofuncRL
- RofuncRL BCQ (Batch-Constrained Q-Learning)
BCQ implementation and explanation of tricks in RofuncRL
- RofuncRL DTrans (Decision Transformer)
DTrans implementation and explanation of tricks in RofuncRL
- RofuncRL TD3+BC (Twin Delayed Deep Deterministic Policy Gradient with Batch-Constrained)
TD3+BC implementation and explanation of tricks in RofuncRL
- RofuncRL EDAC (Ensemble-Diversified Actor Critic)
EDAC implementation and explanation of tricks in RofuncRL
3. Mixline (Mixing online and offline) reinforcement learning#
- RofuncRL AMP (Adversarial Motion Priors)
AMP implementation and explanation of tricks in RofuncRL
- RofuncRL ASE (Adversarial Skill Embeddings)
ASE implementation and explanation of tricks in RofuncRL
- RofuncRL ODTrans (Online Decision Transformer)
ODTrans implementation and explanation of tricks in RofuncRL