Docs: overhaul Getting Started and navigation structure by cvolkcvolk · Pull Request #505 · isaac-sim/IsaacLab-Arena

cvolkcvolk · 2026-03-25T12:55:07Z

Summary

New Getting Started docs: Added first_arena_env.rst walking through the pick_and_place_maple_table environment source with step-by-step explanations of Scene, Embodiment, Task, Relations, and CLI variation. Added first_experiments.rst (formerly Getting Started) with policy runner experiments using zero-action policy.
Renamed srl_pick_and_place → pick_and_place_maple_table: Updated environment class, file, name string, cli.py, all eval job configs, and docs.
Restructured navigation:
- Split User's Guide into Set Up (Installation only — Docker Containers merged in) and Getting Started (First Arena Environment, Running your First Experiments)
- Restructured Example Workflows into Imitation Learning (G1, GR1 Open Microwave, GR1 Sequential) and Reinforcement Learning (Franka Lift Object)
- Moved Jupyter Notebooks and GR00T Server-Client to Advanced
GR00T container decoupled from core setup: Removed Base + GR00T container from Docker Containers setup page. Added a dedicated GR00T Container section to the Imitation Learning index where it is contextually relevant.
Merged Docker Containers into Installation: The page was slim enough after cleanup to fold in directly, simplifying the Set Up section to a single page.

Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Rename the environment class, file, and name string from srl_pick_and_place to pick_and_place_maple_table. Update all references in cli.py, eval job configs, and docs. Also rework first_arena_env.rst to walk through the pick_and_place_maple_table source with step comments, reorder docs so first_arena_env comes before getting_started, and clean up forward references in the intro. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Extract Server-Client Mode With GR00T section from docker_containers.rst into a new advanced/groot_server_client.rst page. Move jupyter_notebooks from User's Guide to Advanced in the ToC. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Add imitation_learning/index.rst grouping the G1 and two GR1 tasks, and reinforcement_learning_workflows/index.rst grouping the Franka Lift Object task. Update main ToC to show these two sections instead of four flat entries. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Split User's Guide toctree into Set Up (Installation, Docker Containers) and Getting Started (First Arena Environment, Varying your Environment). Rename getting_started.rst to varying_your_environment.rst and update its title and all cross-references accordingly. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Remove the GR00T-specific requirements paragraph from Installation and add it as a note on the Imitation Learning index, where it is relevant to the reader. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Document that `python` is aliased to `/isaac-sim/python.sh` inside the container. Recommend using the full path only for docker exec commands run from outside, where the alias is not active. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

first_arena_env now precedes this page in the nav, so pointing readers back to it from Next Steps is confusing. Remove the line. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

GR00T is an external model independent of Arena. The Base + GR00T container is an implementation detail of specific workflows, not part of Arena's core setup. Remove it from Docker Containers and leave it to the Imitation Learning workflow pages to introduce it in context. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Introduce the Base + GR00T container in context, with the launch command and a note that workflow pages indicate when it is needed. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Docker Containers is now slim enough to fold into Installation directly. Move the container launch command and mounted directories note into Installation and remove the separate page and its ToC entry. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Update all workflow step pages that referenced the now-deleted docker_containers page. Base container references now point to installation; Base + GR00T references point to the imitation learning index where the GR00T container is documented. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Rename the page and file to better reflect its purpose as an entry point for hands-on experimentation with Arena. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Signed-off-by: Clemens Volk <cvolk@nvidia.com>

The policy_runner GR00T example now explicitly states it requires the GR00T container (./docker/run_docker.sh -g) before the code block, preventing a silent failure for readers who haven't launched it. Moved the redundant note from the trailing sentence. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Replace GR00T-specific commands with a short pointer to the Imitation Learning workflows and policy evaluation docs. The detailed DROID/GR00T example will live in the workflow section. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Signed-off-by: Clemens Volk <cvolk@nvidia.com>

cvolkcvolk · 2026-03-25T13:52:15Z

docs/pages/quickstart/first_experiments.rst

+----------
+
+The experiments above use ``zero_action`` — the robot stays still and success rates are zero.
+To get non-zero results, swap in a real policy. See :doc:`../../pages/example_workflows/imitation_learning/index`


I feel like the next logical step would be to showcase the Gr00t-Droid zero shot experiment (running in the gr00t container) on this pick_and_place_maple_table environment (Instead of pointing to the imitation learning workflows). There is still quite a big gap in my opinion from zero_action above to the full blown imitation learning pipeline.
Later we would update this to use a simple server-client model and then add a Bring-Your-Model.

cvolkcvolk · 2026-03-25T16:37:46Z

docs/pages/quickstart/first_arena_env.rst

-See :doc:`../concepts/concept_environment_compilation` for how compilation works.
+``IsaacLabArenaEnvironment`` holds the three pieces together.
+This is Arena's declarative description of the environment.
+Simple huh?


xyao-nv · 2026-03-25T17:09:43Z

docs/pages/example_workflows/imitation_learning/index.rst

+Not every step requires this container — the workflow pages will tell you when to use it.
+
+.. note::
+   The Base + GR00T container does not support Blackwell GPUs and requires large amounts of GPU


It does. I will apply the fix directly

xyao-nv · 2026-03-25T17:13:58Z

docs/pages/advanced/groot_server_client.rst

+    --policy_config_yaml_path {policy_config_yaml_path} # e.g. isaaclab_arena_gr00t/policy/config/gr1_manip_gr00t_closedloop_config.yaml
+
+3. Inside the Base container, run the evaluation script with a
+   client-side remote policy (refer to :doc:`../example_workflows/static_manipulation/step_5_evaluation` for full command lines).


This is will not show in SQA version. So Im noting things down loud here.

xyao-nv · 2026-03-25T17:18:44Z

docs/pages/advanced/groot_server_client.rst

+3. Inside the Base container, run the evaluation script with a
+   client-side remote policy (refer to :doc:`../example_workflows/static_manipulation/step_5_evaluation` for full command lines).
+
+This setup cleanly separates the Isaac Lab Arena simulation environment from the GR00T policy


I prefer explaining what 2 containers do exactly than calling it 'seperates'. Will apply modifications

xyao-nv · 2026-03-25T17:36:58Z

docs/pages/quickstart/first_arena_env.rst

+                  object_type=ObjectType.RIGID,
+              )
+
+              # Step 2: Describe spatial relationships


Where is the concept section explaining how spatial relationship works 👀 @cvolkcvolk

xyao-nv · 2026-03-25T17:43:23Z

isaaclab_arena/cli/isaaclab_arena_cli.py

        help="Disable solving spatial relations in the environment.",
    )
+    arena_group.add_argument(
+        "--placement_seed",


Not for now, but for v0.3, I believe we shall pass a handle to all the seed, as env, sim, policy (torch, torch.cuda, rand, np) all use the same seed and now they are all set thru --seed cli.

alexmillane

Massive improvement to our docs. Thanks so much for doing this.

I made some small wording changes as I reviewed and smashed them directly to your branch.

cvolkcvolk added 17 commits March 24, 2026 17:21

Add Getting Started docs with srl_pick_and_place showcase

ba85d70

Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Rework Getting Started page to use policy_runner commands

7c863ed

Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Move GR00T fine-tuning requirements note to Imitation Learning

65bee4e

Remove the GR00T-specific requirements paragraph from Installation and add it as a note on the Imitation Learning index, where it is relevant to the reader. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Clarify python alias in AGENTS.md

dcc24d3

Document that `python` is aliased to `/isaac-sim/python.sh` inside the container. Recommend using the full path only for docker exec commands run from outside, where the alias is not active. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Remove stale back-reference to first_arena_env

86932c5

first_arena_env now precedes this page in the nav, so pointing readers back to it from Next Steps is confusing. Remove the line. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Add GR00T container section to Imitation Learning index

ed3ff31

Introduce the Base + GR00T container in context, with the launch command and a note that workflow pages indicate when it is needed. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Rename Varying your Environment to Running your First Experiments

a4324d0

Rename the page and file to better reflect its purpose as an entry point for hands-on experimentation with Arena. Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Merge branch 'main' into cvolk/getting-started-docs

9694679

Change landing page title to Welcome to Isaac Lab Arena!

417ef81

Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Fix broken grammar in landing page opening sentence

953352b

Signed-off-by: Clemens Volk <cvolk@nvidia.com>

cvolkcvolk changed the base branch from release/0.1.1 to main March 25, 2026 13:01

cvolkcvolk added 4 commits March 25, 2026 14:10

Remove policy evaluation link from Next Steps

6aa183b

Signed-off-by: Clemens Volk <cvolk@nvidia.com>

Remove gif job configs (parked on cvolk/gif-job-configs)

f38674c

Signed-off-by: Clemens Volk <cvolk@nvidia.com>

cvolkcvolk commented Mar 25, 2026

View reviewed changes

alexmillane and others added 2 commits March 25, 2026 16:50

Alex review

602f6f0

Add batch evaluation gif

a7a7121

cvolkcvolk commented Mar 25, 2026

View reviewed changes

xyao-nv reviewed Mar 25, 2026

View reviewed changes

xyao-nv added 2 commits March 25, 2026 10:44

apply xinjie's comments

022dc6b

rename gr00t to gr00t

294bb7c

alexmillane approved these changes Mar 25, 2026

View reviewed changes

xyao-nv approved these changes Mar 25, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Docs: overhaul Getting Started and navigation structure#505

Docs: overhaul Getting Started and navigation structure#505
cvolkcvolk wants to merge 25 commits intomainfrom
cvolk/getting-started-docs

cvolkcvolk commented Mar 25, 2026 •

edited

Loading

Uh oh!

cvolkcvolk Mar 25, 2026

Uh oh!

cvolkcvolk Mar 25, 2026

Uh oh!

xyao-nv Mar 25, 2026

Uh oh!

xyao-nv Mar 25, 2026

Uh oh!

xyao-nv Mar 25, 2026 •

edited

Loading

Uh oh!

xyao-nv Mar 25, 2026

Uh oh!

xyao-nv Mar 25, 2026

Uh oh!

alexmillane left a comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

cvolkcvolk commented Mar 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Uh oh!

cvolkcvolk Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

cvolkcvolk Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

xyao-nv Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

xyao-nv Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

xyao-nv Mar 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

xyao-nv Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

xyao-nv Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

alexmillane left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

cvolkcvolk commented Mar 25, 2026 •

edited

Loading

xyao-nv Mar 25, 2026 •

edited

Loading