# IROS-2025-Challenge-Manip

From the catalogue. Listed by one researcher from one source page.

- id: `iros-2025-challenge-manip`
- kind: synthetic or generated
- robot or device: simulated manipulators (class: simulation)
- how collected: simulation generated pick and place
- size as stated: ~20,000 pretrain trajectories plus ~100 per task
- hours: not stated | hours only claimed: none | episodes: 20,000
- year: 2025
- organisation: Shanghai AI Laboratory (InternRobotics)
- licence: cc-by-nc-sa-4.0 (class: non commercial)
- access: open (class: open)
- link (dataset): https://huggingface.co/datasets/InternRobotics/IROS-2025-Challenge-Manip
- Hugging Face repository: `InternRobotics/IROS-2025-Challenge-Manip`, 15,458 downloads in 30 days, https://huggingface.co/datasets/InternRobotics/IROS-2025-Challenge-Manip
- note: Challenge benchmark with pretrain, train and validation splits.
- download links of one episode: https://datasets.gurasees.com/episodes/InternRobotics/IROS-2025-Challenge-Manip.md?episode=0

Confirm the licence at the link before relying on it.