🐙 Octo: An Open-Source Generalist Robot Policy

Submit
🐙 Octo: An Open-Source Generalist Robot Policygithub.io
github.io

🐙 Octo: An Open-Source Generalist Robot Policy

AI summary of the linked article

Octo is an open-source, transformer-based diffusion policy for robotic manipulation, pretrained on 800k robot episodes from the Open X-Embodiment dataset and designed to be quickly finetuned to new robot setups. The project introduces two initial versions, Octo-Small (27M parameters) and Octo-Base (93M parameters). Out of the box, Octo outperforms RT-1-X when tasks are specified in natural language and performs similarly to the 55-billion-parameter RT-2-X. On the WidowX tasks, goal image conditioning achieved 25% higher performance on average. Across six finetuning evaluation setups, Octo outperforms the next best baseline by 52%, using about 100 target demonstrations per task.

Recommended by 2 curators
No comments yetWhat caught your eye? Start the discussion.