The stable-gym package contains modified versions of the Minitaur environments found in the pybullet package. These environments are different because they return a (positive) cost instead of a (negative) reward, making them compatible with stable RL algorithms. Please refer to the original codebase or the article of Tan et al. 2018 on which the Minitaur environment is based for more information. The stable-gym package currently contains the following MinitaurCost environments:
- MinitaurBullet-v1: The minitaur has to track a reference velocity.