Question about simulation data in implement SPD controller, deepmimic reward functions

Hi, guys, I am trying to implement deepmimic in issacgym in reference of the code in (GitHub - Khrylx/RFC: Official PyTorch Implementation of "Residual Force Control for Agile Human Behavior Imitation and Extended Motion Synthesis". NeurIPS 2020.), which is implemented in mujoco.
Mujoco has provided many data APIs in class mjData, I would like ask how to get some comparable data in mjData from issacgym.

1.while implementing SPD controller, the mjData.frpc_bias and the mjData.qM, where mjData.frpc_bias is compute by C(qpos,qvel) and mjData.qM is the total inertia (from MuJoCo API Reference. or has anyone implemented SPD controller in isaacgym?

  1. while implementing deepmimic reward functions,themjData.subtree_com which is the center of mass of the humanoid.

BTW,Gavriel mentioned in the introduction video that code of AMP in isaacgym might be released. Has it been released, or is there any estimated time when it would be released?

thanks!! I hope I could share my code when I solve these problems.