tobiassteidle/Reinforcement-Learning

Solutions for different Reinforcement Learning environments

Reinforcement-Learning

Repository for OpenAI and Unity-ML Reinforcement Learning environments.

Q - Learning

Taxi-v2 (Solution)

Q - Learning or SARSA

Taxi-v3 (Solution) (Decay ε Greedy)

Pytorch

LunarLanderContinuous-v2 (Solution DDPG)
MountainCarContinuous-v0 (Solution DDPG)
BipedalWalker-v2 (Solution TD3) solved after 1635 episodes
BipedalWalker-v3 (Solution TD3) solved after 678 episodes
HumanoidPyBulletEnv-v0 (Solution PPO)

Tensorflow / Keras

CartPole-v0 (Solution DDQN, Duelling DQN (incl. Prioritized Replay Buffer (PER))