Google DeepMind's software stack for physics-based simulation and Reinforcement Learning environments, using MuJoCo.
In commit https://github.com/deepmind/dm_control/commit/ac6d2cd7af7f6d20bbc0e51df8ba41016a07f1f9, we sped up the warning checks that occur at each physics step. Since these checks are called very often, the performance boost can be considerable in many use cases.
This release is identical to 1.0.3 other than in setup.py.