[HN Gopher] MuJoco Playground
___________________________________________________________________
MuJoco Playground
Author : kzakka
Score : 74 points
Date : 2025-01-16 22:08 UTC (1 days ago)
(HTM) web link (playground.mujoco.org)
(TXT) w3m dump (playground.mujoco.org)
| kzakka wrote:
| An open-source library for GPU-accelerated robot learning and
| sim-to-real transfer.
| archiv wrote:
| This is great - I haven't looked at the documentation yet but do
| you support soft body simulation?
| zone411 wrote:
| Playing with MuJoCo is a lot of fun. I recommend it for people
| who just want to experiment with RL. You can do random weird
| stuff like this: https://www.youtube.com/watch?v=e-9u_vepfY8. I'm
| really glad DeepMind is continuing to update it.
| whatever1 wrote:
| Holy Molly. The live demo where you put a place holder and the
| trained robot dog just goes there is insane! You can also apply
| force (aka kick) the robot while it is moving!
| gunalx wrote:
| Really cool. If we could do sim training to rl conversion in
| robotics, we migth get a llm moment for robotics, with good
| pretrained movement algorithms.
___________________________________________________________________
(page generated 2025-01-17 23:02 UTC)