Skip to content

Additional nit changes to make OmniReset docs more clear #21

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
patrickhaoy merged 1 commit into from
Jan 5, 2026
Merged

Additional nit changes to make OmniReset docs more clear #21

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
108 changes: 102 additions & 6 deletions docs/source/publications/omnireset/index.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -182,6 +182,10 @@ Reproduce our training results from scratch.

.. tab-item:: Leg Twisting

.. note::

**Skip directly to Step 4** if you want to train an RL policy with our pre-generated reset state datasets. Only run Steps 1-3 if you want to generate your own.

**Step 1: Collect Partial Assemblies** (~30 seconds)

.. code:: bash
Expand All @@ -198,7 +202,7 @@ Reproduce our training results from scratch.

.. important::

Before running, make sure ``base_path`` and ``base_paths`` in ``reset_states_cfg.py`` are set appropriately.
Before running, make sure ``base_path`` and ``base_paths`` in ``reset_states_cfg.py`` point to your dataset directories.

.. code:: bash

Expand All @@ -214,8 +218,20 @@ Reproduce our training results from scratch.
# Object Partially Assembled, End-Effector Grasped (Near Goal)
python scripts_v2/tools/record_reset_states.py --task OmniReset-UR5eRobotiq2f85-ObjectPartiallyAssembledEEGrasped-v0 --num_envs 4096 --num_reset_states 10000 --headless --dataset_dir ./reset_state_datasets/ObjectPartiallyAssembledEEGrasped env.scene.insertive_object=fbleg env.scene.receptive_object=fbtabletop

**Step 3.5: Visualize Reset States (Optional)**

Visualize the generated reset states to verify they are correct.

.. code:: bash

python scripts_v2/tools/visualize_reset_states.py --task OmniReset-Ur5eRobotiq2f85-RelCartesianOSC-State-Play-v0 --num_envs 4 --dataset_dir ./reset_state_datasets env.scene.insertive_object=fbleg env.scene.receptive_object=fbtabletop

**Step 4: Train RL Policy**

.. important::

If you generated your own datasets in Steps 1-3, make sure to update ``base_paths`` in ``rl_state_cfg.py`` to point to your dataset directories.

.. code:: bash

python -m torch.distributed.run \
Expand All @@ -232,6 +248,10 @@ Reproduce our training results from scratch.

.. tab-item:: Drawer Assembly

.. note::

**Skip directly to Step 4** if you want to train an RL policy with our pre-generated reset state datasets. Only run Steps 1-3 if you want to generate your own.

**Step 1: Collect Partial Assemblies** (~30 seconds)

.. code:: bash
Expand All @@ -248,7 +268,7 @@ Reproduce our training results from scratch.

.. important::

Before running, make sure ``base_path`` and ``base_paths`` in ``reset_states_cfg.py`` are set appropriately.
Before running, make sure ``base_path`` and ``base_paths`` in ``reset_states_cfg.py`` point to your dataset directories.

.. code:: bash

Expand All @@ -264,8 +284,20 @@ Reproduce our training results from scratch.
# Object Partially Assembled, End-Effector Grasped (Near Goal)
python scripts_v2/tools/record_reset_states.py --task OmniReset-UR5eRobotiq2f85-ObjectPartiallyAssembledEEGrasped-v0 --num_envs 4096 --num_reset_states 10000 --headless --dataset_dir ./reset_state_datasets/ObjectPartiallyAssembledEEGrasped env.scene.insertive_object=fbdrawerbottom env.scene.receptive_object=fbdrawerbox

**Step 3.5: Visualize Reset States (Optional)**

Visualize the generated reset states to verify they are correct.

.. code:: bash

python scripts_v2/tools/visualize_reset_states.py --task OmniReset-Ur5eRobotiq2f85-RelCartesianOSC-State-Play-v0 --num_envs 4 --dataset_dir ./reset_state_datasets env.scene.insertive_object=fbdrawerbottom env.scene.receptive_object=fbdrawerbox

**Step 4: Train RL Policy**

.. important::

If you generated your own datasets in Steps 1-3, make sure to update ``base_paths`` in ``rl_state_cfg.py`` to point to your dataset directories.

.. code:: bash

python -m torch.distributed.run \
Expand All @@ -282,6 +314,10 @@ Reproduce our training results from scratch.

.. tab-item:: Peg Insertion

.. note::

**Skip directly to Step 4** if you want to train an RL policy with our pre-generated reset state datasets. Only run Steps 1-3 if you want to generate your own.

**Step 1: Collect Partial Assemblies** (~30 seconds)

.. code:: bash
Expand All @@ -298,7 +334,7 @@ Reproduce our training results from scratch.

.. important::

Before running, make sure ``base_path`` and ``base_paths`` in ``reset_states_cfg.py`` are set appropriately.
Before running, make sure ``base_path`` and ``base_paths`` in ``reset_states_cfg.py`` point to your dataset directories.

.. code:: bash

Expand All @@ -314,8 +350,20 @@ Reproduce our training results from scratch.
# Object Partially Assembled, End-Effector Grasped (Near Goal)
python scripts_v2/tools/record_reset_states.py --task OmniReset-UR5eRobotiq2f85-ObjectPartiallyAssembledEEGrasped-v0 --num_envs 4096 --num_reset_states 10000 --headless --dataset_dir ./reset_state_datasets/ObjectPartiallyAssembledEEGrasped env.scene.insertive_object=peg env.scene.receptive_object=peghole

**Step 3.5: Visualize Reset States (Optional)**

Visualize the generated reset states to verify they are correct.

.. code:: bash

python scripts_v2/tools/visualize_reset_states.py --task OmniReset-Ur5eRobotiq2f85-RelCartesianOSC-State-Play-v0 --num_envs 4 --dataset_dir ./reset_state_datasets env.scene.insertive_object=peg env.scene.receptive_object=peghole

**Step 4: Train RL Policy**

.. important::

If you generated your own datasets in Steps 1-3, make sure to update ``base_paths`` in ``rl_state_cfg.py`` to point to your dataset directories.

.. code:: bash

python -m torch.distributed.run \
Expand All @@ -332,6 +380,10 @@ Reproduce our training results from scratch.

.. tab-item:: Rectangle on Wall

.. note::

**Skip directly to Step 4** if you want to train an RL policy with our pre-generated reset state datasets. Only run Steps 1-3 if you want to generate your own.

**Step 1: Collect Partial Assemblies** (~30 seconds)

.. code:: bash
Expand All @@ -348,7 +400,7 @@ Reproduce our training results from scratch.

.. important::

Before running, make sure ``base_path`` and ``base_paths`` in ``reset_states_cfg.py`` are set appropriately.
Before running, make sure ``base_path`` and ``base_paths`` in ``reset_states_cfg.py`` point to your dataset directories.

.. code:: bash

Expand All @@ -364,8 +416,20 @@ Reproduce our training results from scratch.
# Object Partially Assembled, End-Effector Grasped (Near Goal)
python scripts_v2/tools/record_reset_states.py --task OmniReset-UR5eRobotiq2f85-ObjectPartiallyAssembledEEGrasped-v0 --num_envs 4096 --num_reset_states 10000 --headless --dataset_dir ./reset_state_datasets/ObjectPartiallyAssembledEEGrasped env.scene.insertive_object=rectangle env.scene.receptive_object=wall

**Step 3.5: Visualize Reset States (Optional)**

Visualize the generated reset states to verify they are correct.

.. code:: bash

python scripts_v2/tools/visualize_reset_states.py --task OmniReset-Ur5eRobotiq2f85-RelCartesianOSC-State-Play-v0 --num_envs 4 --dataset_dir ./reset_state_datasets env.scene.insertive_object=rectangle env.scene.receptive_object=wall

**Step 4: Train RL Policy**

.. important::

If you generated your own datasets in Steps 1-3, make sure to update ``base_paths`` in ``rl_state_cfg.py`` to point to your dataset directories.

.. code:: bash

python -m torch.distributed.run \
Expand All @@ -382,6 +446,10 @@ Reproduce our training results from scratch.

.. tab-item:: Cube Stacking

.. note::

**Skip directly to Step 4** if you want to train an RL policy with our pre-generated reset state datasets. Only run Steps 1-3 if you want to generate your own.

**Step 1: Collect Partial Assemblies** (~30 seconds)

.. code:: bash
Expand All @@ -398,7 +466,7 @@ Reproduce our training results from scratch.

.. important::

Before running, make sure ``base_path`` and ``base_paths`` in ``reset_states_cfg.py`` are set appropriately.
Before running, make sure ``base_path`` and ``base_paths`` in ``reset_states_cfg.py`` point to your dataset directories.

.. code:: bash

Expand All @@ -414,8 +482,20 @@ Reproduce our training results from scratch.
# Object Partially Assembled, End-Effector Grasped (Near Goal)
python scripts_v2/tools/record_reset_states.py --task OmniReset-UR5eRobotiq2f85-ObjectPartiallyAssembledEEGrasped-v0 --num_envs 4096 --num_reset_states 10000 --headless --dataset_dir ./reset_state_datasets/ObjectPartiallyAssembledEEGrasped env.scene.insertive_object=cube env.scene.receptive_object=cube

**Step 3.5: Visualize Reset States (Optional)**

Visualize the generated reset states to verify they are correct.

.. code:: bash

python scripts_v2/tools/visualize_reset_states.py --task OmniReset-Ur5eRobotiq2f85-RelCartesianOSC-State-Play-v0 --num_envs 4 --dataset_dir ./reset_state_datasets env.scene.insertive_object=cube env.scene.receptive_object=cube

**Step 4: Train RL Policy**

.. important::

If you generated your own datasets in Steps 1-3, make sure to update ``base_paths`` in ``rl_state_cfg.py`` to point to your dataset directories.

.. code:: bash

python -m torch.distributed.run \
Expand All @@ -432,6 +512,10 @@ Reproduce our training results from scratch.

.. tab-item:: Cupcake on Plate

.. note::

**Skip directly to Step 4** if you want to train an RL policy with our pre-generated reset state datasets. Only run Steps 1-3 if you want to generate your own.

**Step 1: Collect Partial Assemblies** (~30 seconds)

.. code:: bash
Expand All @@ -448,7 +532,7 @@ Reproduce our training results from scratch.

.. important::

Before running, make sure ``base_path`` and ``base_paths`` in ``reset_states_cfg.py`` are set appropriately.
Before running, make sure ``base_path`` and ``base_paths`` in ``reset_states_cfg.py`` point to your dataset directories.

.. code:: bash

Expand All @@ -464,8 +548,20 @@ Reproduce our training results from scratch.
# Object Partially Assembled, End-Effector Grasped (Near Goal)
python scripts_v2/tools/record_reset_states.py --task OmniReset-UR5eRobotiq2f85-ObjectPartiallyAssembledEEGrasped-v0 --num_envs 4096 --num_reset_states 10000 --headless --dataset_dir ./reset_state_datasets/ObjectPartiallyAssembledEEGrasped env.scene.insertive_object=cupcake env.scene.receptive_object=plate

**Step 3.5: Visualize Reset States (Optional)**

Visualize the generated reset states to verify they are correct.

.. code:: bash

python scripts_v2/tools/visualize_reset_states.py --task OmniReset-Ur5eRobotiq2f85-RelCartesianOSC-State-Play-v0 --num_envs 4 --dataset_dir ./reset_state_datasets env.scene.insertive_object=cupcake env.scene.receptive_object=plate

**Step 4: Train RL Policy**

.. important::

If you generated your own datasets in Steps 1-3, make sure to update ``base_paths`` in ``rl_state_cfg.py`` to point to your dataset directories.

.. code:: bash

python -m torch.distributed.run \
Expand Down
Loading