# Soft Actor Critic Each experiment uses 3 seeds and is trained for 3M environment steps. The parameters used for SAC are the same parameters as described in the [original paper](https://arxiv.org/abs/1801.01290). ### Inverted Pendulum SAC - single worker ```bash coach -p Mujoco_SAC -lvl inverted_pendulum ``` Inverted Pendulum SAC ### Hopper Clipped SAC - single worker ```bash coach -p Mujoco_SAC -lvl hopper ``` Hopper SAC ### Half Cheetah Clipped SAC - single worker ```bash coach -p Mujoco_SAC -lvl half_cheetah ``` Half Cheetah SAC ### Walker 2D Clipped SAC - single worker ```bash coach -p Mujoco_SAC -lvl walker2d ``` Walker 2D SAC ### Humanoid Clipped SAC - single worker ```bash coach -p Mujoco_SAC -lvl humanoid ``` Humanoid SAC