Object_picking / README.md
Sampath-d's picture
Update README.md
6647540 verified
|
Raw History Blame Contribute Delete
1.39 kB
---
license: apache-2.0
task_categories:
- robotics
tags:
- LeRobot
configs:
- config_name: default
data_files: data/*/*.parquet
---
This dataset was created using [LeRobot](https://github.com/huggingface/lerobot).
<a class="flex" href="https://huggingface.co/spaces/lerobot/visualize_dataset?path=Sampath-d/Object_picking">
<img class="block dark:hidden" src="https://huggingface.co/datasets/huggingface/badges/resolve/main/visualize-this-dataset-xl.svg"/>
<img class="hidden dark:block" src="https://huggingface.co/datasets/huggingface/badges/resolve/main/visualize-this-dataset-xl-dark.svg"/>
</a>
# XLe SO-101 Object Picking
Real-world object-picking dataset for **fine-tuning [allenai/MolmoAct2-SO100_101](https://huggingface.co/allenai/MolmoAct2-SO100_101)** on an **SO-101 follower arm mounted on XLeRobot**.
Stored in **LeRobot v3.0** format.
* **Episodes:** 60
* **Frames:** 47,175
* **FPS:** 30
* **Resolution:** 640 × 480 RGB
* **Cameras:** Top + Wrist
* **State / Action:** 6 SO-101 joint positions
* **Tasks:** 4
The dataset contains synchronized camera observations, robot states, and actions collected from real SO-101 demonstrations on XLeRobot.
### Base Checkpoint
[allenai/MolmoAct2-SO100_101](https://huggingface.co/allenai/MolmoAct2-SO100_101)
### Format
LeRobot v3.0 with Parquet data, episode/task metadata, statistics, and encoded camera videos.