Merge branch 'main' into feature/g1imitation

Author: fei-yang-wu

CONTRIBUTORS.md

@@ -40,11 +40,13 @@ Guidelines for modifications:
 * Amr Mousa
 * Andrej Orsula
 * Anton Bjørndahl Mortensen
+* Antonin Raffin
 * Arjun Bhardwaj
 * Ashwin Varghese Kuruttukulam
 * Bikram Pandit
 * Bingjie Tang
 * Brayden Zhang
+* Brian Bingham
 * Cameron Upright
 * Calvin Yu
 * Cheng-Rong Lai
@@ -74,6 +76,7 @@ Guidelines for modifications:
 * Jinqi Wei
 * Johnson Sun
 * Kaixi Bao
+* Kris Wilson
 * Kourosh Darvish
 * Kousheek Chakraborty
 * Lionel Gulich
@@ -86,6 +89,7 @@ Guidelines for modifications:
 * Miguel Alonso Jr
 * Muhong Guo
 * Nicola Loi
+* Norbert Cygiert
 * Nuoyan Chen (Alvin)
 * Nuralem Abizov
 * Ori Gadot
@@ -102,6 +106,7 @@ Guidelines for modifications:
 * Rosario Scalise
 * Ryley McCarroll
 * Shafeef Omar
+* Shaoshu Su
 * Shundo Kishi
 * Stefan Van de Mosselaer
 * Stephan Pleines
@@ -119,6 +124,7 @@ Guidelines for modifications:
 * Zhengyu Zhang
 * Ziqi Fan
 * Zoe McCarthy
+* David Leon
 
 ## Acknowledgements
 

README.md

@@ -97,12 +97,13 @@ or opening a question on its [forums](https://forums.developer.nvidia.com/c/agx-
 
 ## Connect with the NVIDIA Omniverse Community
 
-Have a project or resource you'd like to share more widely? We'd love to hear from you! Reach out to the
-NVIDIA Omniverse Community team at <[email protected]> to discuss potential opportunities
-for broader dissemination of your work.
+Do you have a project or resource you'd like to share more widely? We'd love to hear from you!
+Reach out to the NVIDIA Omniverse Community team at [email protected] to explore opportunities
+to spotlight your work.
 
-Join us in building a vibrant, collaborative ecosystem where creativity and technology intersect. Your
-contributions can make a significant impact on the Isaac Lab community and beyond!
+You can also join the conversation on the [Omniverse Discord](https://discord.com/invite/nvidiaomniverse) to
+connect with other developers, share your projects, and help grow a vibrant, collaborative ecosystem
+where creativity and technology intersect. Your contributions can make a meaningful impact on the Isaac Lab community and beyond!
 
 ## License
 

docs/source/_static/tasks/locomotion/agility_digit_flat.jpg

@@ -63,7 +63,7 @@ Integrating a new library
 
 Adding a new library to Isaac Lab is similar to using a different version of a library. You can install the library
 in your Python environment and use it in your experiments. However, if you want to integrate the library with
-Isaac Lab, you can will first need to make a wrapper for the library, as explained in
+Isaac Lab, you will first need to make a wrapper for the library, as explained in
 :ref:`how-to-env-wrappers`.
 
 The following steps can be followed to integrate a new library with Isaac Lab:

docs/source/_static/tasks/locomotion/agility_digit_loco_manip.jpg

@@ -35,6 +35,7 @@ maximum effort:
 .. math::
 
     \tau_{j, computed} & = k_p * (q_{des} - q) + k_d * (\dot{q}_{des} - \dot{q}) + \tau_{ff} \\
+    \tau_{j, max} & = \gamma \times \tau_{motor, max} \\
     \tau_{j, applied} & = clip(\tau_{computed}, -\tau_{j, max}, \tau_{j, max})
 
 

docs/source/_static/tasks/locomotion/agility_digit_rough.jpg

@@ -214,6 +214,12 @@ We provide environments for both disassembly and assembly.
       wget https://developer.download.nvidia.com/compute/cuda/12.8.0/local_installers/cuda_12.8.0_570.86.10_linux.run
       sudo sh cuda_12.8.0_570.86.10_linux.run
 
+  When using conda, cuda toolkit can be installed with:
+
+  .. code-block:: bash
+
+      conda install cudatoolkit
+
   For addition instructions and Windows installation, please refer to the `CUDA installation page <https://developer.nvidia.com/cuda-12-8-0-download-archive>`_.
 
 * |disassembly-link|: The plug starts inserted in the socket. A low-level controller lifts the plug out and moves it to a random position. This process is purely scripted and does not involve any learned policy. Therefore, it does not require policy training or evaluation. The resulting trajectories serve as demonstrations for the reverse process, i.e., learning to assemble. To run disassembly for a specific task: ``python source/isaaclab_tasks/isaaclab_tasks/direct/automate/run_disassembly_w_id.py --assembly_id=ASSEMBLY_ID --disassembly_dir=DISASSEMBLY_DIR``. All generated trajectories are saved to a local directory ``DISASSEMBLY_DIR``.
@@ -288,6 +294,12 @@ Environments based on legged locomotion tasks.
     +------------------------------+----------------------------------------------+------------------------------------------------------------------------------+
     | |velocity-rough-g1|          | |velocity-rough-g1-link|                     | Track a velocity command on rough terrain with the Unitree G1 robot          |
     +------------------------------+----------------------------------------------+------------------------------------------------------------------------------+
+    | |velocity-flat-digit|        | |velocity-flat-digit-link|                   | Track a velocity command on flat terrain with the Agility Digit robot        |
+    +------------------------------+----------------------------------------------+------------------------------------------------------------------------------+
+    | |velocity-rough-digit|       | |velocity-rough-digit-link|                  | Track a velocity command on rough terrain with the Agility Digit robot       |
+    +------------------------------+----------------------------------------------+------------------------------------------------------------------------------+
+    | |tracking-loco-manip-digit|  | |tracking-loco-manip-digit-link|             | Track a root velocity and hand pose command with the Agility Digit robot     |
+    +------------------------------+----------------------------------------------+------------------------------------------------------------------------------+
 
 .. |velocity-flat-anymal-b-link| replace:: `Isaac-Velocity-Flat-Anymal-B-v0 <https://github.com/isaac-sim/IsaacLab/blob/main/source/isaaclab_tasks/isaaclab_tasks/manager_based/locomotion/velocity/config/anymal_b/flat_env_cfg.py>`__
 .. |velocity-rough-anymal-b-link| replace:: `Isaac-Velocity-Rough-Anymal-B-v0 <https://github.com/isaac-sim/IsaacLab/blob/main/source/isaaclab_tasks/isaaclab_tasks/manager_based/locomotion/velocity/config/anymal_b/rough_env_cfg.py>`__
@@ -318,6 +330,9 @@ Environments based on legged locomotion tasks.
 .. |velocity-flat-g1-link| replace:: `Isaac-Velocity-Flat-G1-v0 <https://github.com/isaac-sim/IsaacLab/blob/main/source/isaaclab_tasks/isaaclab_tasks/manager_based/locomotion/velocity/config/g1/flat_env_cfg.py>`__
 .. |velocity-rough-g1-link| replace:: `Isaac-Velocity-Rough-G1-v0 <https://github.com/isaac-sim/IsaacLab/blob/main/source/isaaclab_tasks/isaaclab_tasks/manager_based/locomotion/velocity/config/g1/rough_env_cfg.py>`__
 
+.. |velocity-flat-digit-link| replace:: `Isaac-Velocity-Flat-Digit-v0 <https://github.com/isaac-sim/IsaacLab/blob/main/source/isaaclab_tasks/isaaclab_tasks/manager_based/locomotion/velocity/config/digit/flat_env_cfg.py>`__
+.. |velocity-rough-digit-link| replace:: `Isaac-Velocity-Rough-Digit-v0 <https://github.com/isaac-sim/IsaacLab/blob/main/source/isaaclab_tasks/isaaclab_tasks/manager_based/locomotion/velocity/config/digit/rough_env_cfg.py>`__
+.. |tracking-loco-manip-digit-link| replace:: `Isaac-Tracking-Flat-Digit-v0 <https://github.com/isaac-sim/IsaacLab/blob/main/source/isaaclab_tasks/isaaclab_tasks/manager_based/locomotion/tracking/config/digit/loco_manip_env_cfg.py>`__
 
 .. |velocity-flat-anymal-b| image:: ../_static/tasks/locomotion/anymal_b_flat.jpg
 .. |velocity-rough-anymal-b| image:: ../_static/tasks/locomotion/anymal_b_rough.jpg
@@ -336,6 +351,9 @@ Environments based on legged locomotion tasks.
 .. |velocity-rough-h1| image:: ../_static/tasks/locomotion/h1_rough.jpg
 .. |velocity-flat-g1| image:: ../_static/tasks/locomotion/g1_flat.jpg
 .. |velocity-rough-g1| image:: ../_static/tasks/locomotion/g1_rough.jpg
+.. |velocity-flat-digit| image:: ../_static/tasks/locomotion/agility_digit_flat.jpg
+.. |velocity-rough-digit| image:: ../_static/tasks/locomotion/agility_digit_rough.jpg
+.. |tracking-loco-manip-digit| image:: ../_static/tasks/locomotion/agility_digit_loco_manip.jpg
 
 Navigation
 ~~~~~~~~~~
@@ -554,6 +572,10 @@ Environments based on fixed-arm manipulation tasks.
 Comprehensive List of Environments
 ==================================
 
+For environments that have a different task name listed under ``Inference Task Name``, please use the Inference Task Name
+provided when running ``play.py`` or any inferencing workflows. These tasks provide more suitable configurations for
+inferencing, including reading from an already trained checkpoint and disabling runtime perturbations used for training.
+
 .. list-table::
     :widths: 33 25 19 25
 
@@ -761,6 +783,10 @@ Comprehensive List of Environments
       -
       - Manager Based
       -
+    * - Isaac-Tracking-LocoManip-Digit-v0
+      - Isaac-Tracking-LocoManip-Digit-Play-v0
+      - Manager Based
+      - **rsl_rl** (PPO)
     * - Isaac-Navigation-Flat-Anymal-C-v0
       - Isaac-Navigation-Flat-Anymal-C-Play-v0
       - Manager Based
@@ -869,6 +895,10 @@ Comprehensive List of Environments
       - Isaac-Velocity-Flat-Cassie-Play-v0
       - Manager Based
       - **rsl_rl** (PPO), **skrl** (PPO)
+    * - Isaac-Velocity-Flat-Digit-v0
+      - Isaac-Velocity-Flat-Digit-Play-v0
+      - Manager Based
+      - **rsl_rl** (PPO)
     * - Isaac-Velocity-Flat-G1-v0
       - Isaac-Velocity-Flat-G1-Play-v0
       - Manager Based
@@ -884,7 +914,7 @@ Comprehensive List of Environments
     * - Isaac-Velocity-Flat-Unitree-A1-v0
       - Isaac-Velocity-Flat-Unitree-A1-Play-v0
       - Manager Based
-      - **rsl_rl** (PPO), **skrl** (PPO)
+      - **rsl_rl** (PPO), **skrl** (PPO), **sb3** (PPO)
     * - Isaac-Velocity-Flat-Unitree-Go1-v0
       - Isaac-Velocity-Flat-Unitree-Go1-Play-v0
       - Manager Based
@@ -913,6 +943,10 @@ Comprehensive List of Environments
       - Isaac-Velocity-Rough-Cassie-Play-v0
       - Manager Based
       - **rsl_rl** (PPO), **skrl** (PPO)
+    * - Isaac-Velocity-Rough-Digit-v0
+      - Isaac-Velocity-Rough-Digit-Play-v0
+      - Manager Based
+      - **rsl_rl** (PPO)
     * - Isaac-Velocity-Rough-G1-v0
       - Isaac-Velocity-Rough-G1-Play-v0
       - Manager Based
@@ -924,7 +958,7 @@ Comprehensive List of Environments
     * - Isaac-Velocity-Rough-Unitree-A1-v0
       - Isaac-Velocity-Rough-Unitree-A1-Play-v0
       - Manager Based
-      - **rsl_rl** (PPO), **skrl** (PPO)
+      - **rsl_rl** (PPO), **skrl** (PPO), **sb3** (PPO)
     * - Isaac-Velocity-Rough-Unitree-Go1-v0
       - Isaac-Velocity-Rough-Unitree-Go1-Play-v0
       - Manager Based

docs/source/how-to/add_own_library.rst

@@ -187,7 +187,7 @@ Stable-Baselines3
 
 -  Training an agent with
    `Stable-Baselines3 <https://stable-baselines3.readthedocs.io/en/master/index.html>`__
-   on ``Isaac-Cartpole-v0``:
+   on ``Isaac-Velocity-Flat-Unitree-A1-v0``:
 
    .. tab-set::
       :sync-group: os
@@ -200,14 +200,13 @@ Stable-Baselines3
             # install python module (for stable-baselines3)
             ./isaaclab.sh -i sb3
             # run script for training
-            # note: we set the device to cpu since SB3 doesn't optimize for GPU anyway
-            ./isaaclab.sh -p scripts/reinforcement_learning/sb3/train.py --task Isaac-Cartpole-v0 --headless --device cpu
+            ./isaaclab.sh -p scripts/reinforcement_learning/sb3/train.py --task Isaac-Velocity-Flat-Unitree-A1-v0 --headless
             # run script for playing with 32 environments
-            ./isaaclab.sh -p scripts/reinforcement_learning/sb3/play.py --task Isaac-Cartpole-v0 --num_envs 32 --checkpoint /PATH/TO/model.zip
+            ./isaaclab.sh -p scripts/reinforcement_learning/sb3/play.py --task Isaac-Velocity-Flat-Unitree-A1-v0 --num_envs 32 --checkpoint /PATH/TO/model.zip
             # run script for playing a pre-trained checkpoint with 32 environments
-            ./isaaclab.sh -p scripts/reinforcement_learning/sb3/play.py --task Isaac-Cartpole-v0 --num_envs 32 --use_pretrained_checkpoint
+            ./isaaclab.sh -p scripts/reinforcement_learning/sb3/play.py --task Isaac-Velocity-Flat-Unitree-A1-v0 --num_envs 32 --use_pretrained_checkpoint
             # run script for recording video of a trained agent (requires installing `ffmpeg`)
-            ./isaaclab.sh -p scripts/reinforcement_learning/sb3/play.py --task Isaac-Cartpole-v0 --headless --video --video_length 200
+            ./isaaclab.sh -p scripts/reinforcement_learning/sb3/play.py --task Isaac-Velocity-Flat-Unitree-A1-v0 --headless --video --video_length 200
 
       .. tab-item:: :icon:`fa-brands fa-windows` Windows
          :sync: windows
@@ -217,14 +216,13 @@ Stable-Baselines3
             :: install python module (for stable-baselines3)
             isaaclab.bat -i sb3
             :: run script for training
-            :: note: we set the device to cpu since SB3 doesn't optimize for GPU anyway
-            isaaclab.bat -p scripts\reinforcement_learning\sb3\train.py --task Isaac-Cartpole-v0 --headless --device cpu
+            isaaclab.bat -p scripts\reinforcement_learning\sb3\train.py --task Isaac-Velocity-Flat-Unitree-A1-v0 --headless
             :: run script for playing with 32 environments
-            isaaclab.bat -p scripts\reinforcement_learning\sb3\play.py --task Isaac-Cartpole-v0 --num_envs 32 --checkpoint /PATH/TO/model.zip
+            isaaclab.bat -p scripts\reinforcement_learning\sb3\play.py --task Isaac-Velocity-Flat-Unitree-A1-v0 --num_envs 32 --checkpoint /PATH/TO/model.zip
             :: run script for playing a pre-trained checkpoint with 32 environments
-            isaaclab.bat -p scripts\reinforcement_learning\sb3\play.py --task Isaac-Cartpole-v0 --num_envs 32 --use_pretrained_checkpoint
+            isaaclab.bat -p scripts\reinforcement_learning\sb3\play.py --task Isaac-Velocity-Flat-Unitree-A1-v0 --num_envs 32 --use_pretrained_checkpoint
             :: run script for recording video of a trained agent (requires installing `ffmpeg`)
-            isaaclab.bat -p scripts\reinforcement_learning\sb3\play.py --task Isaac-Cartpole-v0 --headless --video --video_length 200
+            isaaclab.bat -p scripts\reinforcement_learning\sb3\play.py --task Isaac-Velocity-Flat-Unitree-A1-v0 --headless --video --video_length 200
 
 All the scripts above log the training progress to `Tensorboard`_ in the ``logs`` directory in the root of
 the repository. The logs directory follows the pattern ``logs/<library>/<task>/<date-time>``, where ``<library>``

docs/source/overview/core-concepts/actuators.rst

@@ -71,17 +71,18 @@ Training Performance
 --------------------
 
 We performed training with each RL library on the same ``Isaac-Humanoid-v0`` environment
-with ``--headless`` on a single RTX 4090 GPU
+with ``--headless`` on a single RTX PRO 6000 GPU using 4096 environments
 and logged the total training time for 65.5M steps for each RL library.
 
+
 +--------------------+-----------------+
 | RL Library         | Time in seconds |
 +====================+=================+
-| RL-Games           | 203             |
+| RL-Games           | 207             |
 +--------------------+-----------------+
-| SKRL               | 204             |
+| SKRL               | 208             |
 +--------------------+-----------------+
-| RSL RL             | 207             |
+| RSL RL             | 199             |
 +--------------------+-----------------+
-| Stable-Baselines3  | 6320            |
+| Stable-Baselines3  | 322             |
 +--------------------+-----------------+