Updates for MuJoCo Warp documentation.

PiperOrigin-RevId: 837027234
Change-Id: Idf7a69b14ce23d5ac5ed36367a14c77a00276846
This commit is contained in:
Taylor Howell
2025-11-26 02:25:24 -08:00
committed by Copybara-Service
parent a9fe8340bf
commit f2823836a6
+1 -4
View File
@@ -256,7 +256,7 @@ Solver iterations
MuJoCo's default solver settings for the maximum numbers of :ref:`solver iterations<option-iterations>` and
:ref:`linesearch iterations<option-ls_iterations>` are expected to provide reasonable performance. Reducing MJWarp's
settings :attr:`Option.iterations <mujoco_warp.Option.iterations>` and/or
:attr:`Optiona.ls_iterations <mujoco_warp.Option.ls_iterations>` limits may improve performance and should be secondary
:attr:`Option.ls_iterations <mujoco_warp.Option.ls_iterations>` limits may improve performance and should be secondary
considerations after tuning `nconmax`_ / `naconmax`_ and `njmax`_.
Reducing these limits too much may prevent the constraint solver from converging and can lead to inaccurate or unstable
@@ -418,9 +418,6 @@ for details and
`mjlab distributed training <https://github.com/mujocolab/mjlab/tree/main/docs/api/distributed_training.md>`__ for a
reinforcement learning example.
Orientation representation
--------------------------
**How are orientations represented?**
Orientations are represented as unit quaternions and follow :ref:`MuJoCo's convention <floating-objects>`: