| license: apache-2.0 | |
| task_categories: | |
| - robotics | |
| tags: | |
| - lerobot | |
| - so101 | |
| - simstudio | |
| - mujoco | |
| - pick-and-place | |
| - teleoperation | |
| size_categories: | |
| - 10K<n<100K | |
| # SO-101 Lab01 Pick-and-Place Dataset | |
| Expert demonstrations for MuJoCo sim2sim pick-and-place, recorded with **[SO-101 SimStudio](https://github.com/rocPAI-Forge/so101-simstudio)** (real leader arm → MuJoCo follower, LeRobot v3.0). | |
| | | | | |
| |---|---| | |
| | Episodes | 50 @ 20 Hz | | |
| | Task | Pick up the cube and place it in the box. | | |
| | Robot | `so101_mujoco` (6-D joint position actions) | | |
| | Cameras | `camera_front`, `camera_top`, `camera_wrist` (640×480) | | |
| ## Documentation | |
| | Resource | Link | | |
| |----------|------| | |
| | SimStudio repo | [rocPAI-Forge/so101-simstudio](https://github.com/rocPAI-Forge/so101-simstudio) | | |
| | Lab 01 (record → validate → train → eval) | [labs/lab01_pnp/lab01_pnp.md](https://github.com/rocPAI-Forge/so101-simstudio/blob/main/labs/lab01_pnp/lab01_pnp.md) | | |
| For recording setup, validation, replay, and training commands, see **Lab 01** in the repo above. | |
| ## Fine-tuned policy | |
| Policies trained on this dataset: | |
| - ACT: [alexhegit/so101-simstudio-lab01-pnp-act](https://huggingface.co/alexhegit/so101-simstudio-lab01-pnp-act) | |
| - SmolVLA (MI300X bs64 @ 50K): [alexhegit/so101-simstudio-lab01-pnp-smolvla](https://huggingface.co/alexhegit/so101-simstudio-lab01-pnp-smolvla) | |
| ## Quick load | |
| ```python | |
| from lerobot.datasets import LeRobotDataset | |
| dataset = LeRobotDataset("alexhegit/so101-simstudio-lab01-pnp") | |
| sample = dataset[0] | |
| ``` | |