EnCortex-Framework

Reinforcement Learning

We currently support PyTorch based Reinforcement Learning frameworks. EnCortex integrates with Stable-Baselines 3, D3RLPY and has a native implementation of some algorithms as well.

Stable Baselines 3

1. DQN
2. A2C
3. PPO

D3RLPY

1. Discrete SAC

Native

1. DQN
2. Noisy DQN

In addition to this, we also support Offline Reinforcement Learning via D3RLPY. To collect a D3RLPY compatible dataset, OfflineCollectTrajectoryCallback can be used directly with any RL environment. Usage: from encortex.callbacks.offline_rl_callback import OfflineCollectTrajectoryCallback

This site is open source. Improve this page.