usability of environment

[seann999]

A couple of comments/issues:

1. The reward and done values seem to be flipped
2. Reward is always 0 and is not indicative of anything; incrementing by 1 for every subgoal achieved would be nice How do I figure out how many subgoals were achieved?

[etaoxing]

1. Thanks for finding that, just pushed a fix (https://github.com/etaoxing/kitchen-shift/commit/6d0928000d4843df43791d3efd9e58af72c3acac), though this shouldn't affect results.

2. Reward and done were computed in a separate wrapper. The o.g adept_envs Env class also just zeroed out rewards and done.

I'd also recommend checking out IsaacGymEnvs; I am no longer actively using this environment / Mujoco.