OpenAI Gym
OpenAI Gym is an open-source toolkit designed to implement a variety of reinforcement learning benchmark tasks, including classic control, Atari games, robotics, and MuJoCo simulation environments. It provides researchers and developers with a standardized platform for testing and comparing different reinforcement learning algorithms, promoting research progress and technological innovation in the field.
Ant-v4
HalfCheetah-v4
SAC
Hopper-v4
Humanoid-v4
Walker2d-v4
Ant-v2
AWR
CartPole-v1
Oblique decision tree
HalfCheetah-v2
TLA
Hopper-v2
AWR
LunarLander-v2
Oblique decision tree
Mountain Car
Pendulum-v1
TLA with Hierarchical Reward Functions
Walker2d-v2
AWR
Humanoid-v2
AWR
InvertedDoublePendulum-v2
InvertedPendulum-v2
MountainCarContinuous-v0