Soroush Nasiriany

Hi, Are you looking to simulate shape shapes (boxes, cylinders, etc) or more complex shapes (a shoe, mug, etc)? For the former case robosuite offers a number of [primitive shapes](https://github.com/ARISE-Initiative/robosuite/tree/master/robosuite/models/objects/primitive)...

I encountered this issue as well, and @kovacgrgur 's solution works! @vitchyr @anair13 would you mind adding this in?

@vitchyr here it is: https://github.com/rail-berkeley/rlkit/pull/142

Thanks for the question. Indeed it is difficult to rely on `mujoco-py` for later versions of Mujoco. We are working right now on removing the dependency on `pip install mujoco-py`...

I tried these steps a while back for Ubuntu, hopefully they end up working for you: * Download `mujoco-2.1.1` into `~/.mujoco` directory * Make a “fake” `mujoco210` folder in the...

Hi @wuphilipp @danijar, thanks for bringing this to our attention. We'll need to do some investigation to understand the difference in speed between robosuite and dm_control. This week is busy...

Hi @yuvaltassa, we really appreciate your efforts in speeding up the simulation! I will share with you some models next week (this week we are all working towards deadlines). Thanks!

Hi @wuphilipp @danijar we've done some profiling of our code, and we found that our controller logic (computing raw low-level actuation) takes up a significant portion of the compute. We...

@yuvaltassa robosuite offers several controllers (IK, joint position control, operational space control, etc). Our most commonly used controller is [operational space control (OSC)](https://studywolf.wordpress.com/2013/09/17/robot-control-4-operation-space-control/). The team will take a closer look...

Thanks @kevinzakka! That did solve it, but it should only be used if the geom is small. In the larger model with 100x scale, I need to make that shellinertia=false....