RACER augmented RLBench
From the catalogue. Listed by one researcher from one source page.
- id:
racer-augmented-rlbench - kind: language and reasoning annotation
- robot or device: Franka Panda in RLBench (class: single arm)
- how collected: perturbed expert demos with GPT-4 generated language
- size as stated: 2250 expert episodes and 10,159 recovery episodes
- hours: not stated | hours only claimed: none | episodes: 10,159
- year: 2024
- organisation: University of Michigan SLED
- licence: mit (class: open)
- access: open (class: open)
- link (dataset): https://huggingface.co/datasets/sled-umich/RACER-augmented_rlbench
- Hugging Face repository:
sled-umich/RACER-augmented_rlbench, 925 downloads in 30 days, https://huggingface.co/datasets/sled-umich/RACER-augmented_rlbench - note: Failure recovery data with rich language.
Confirm the licence at the link before relying on it.