Rubik's cube ADR robot

Training compute

8.5×10²⁰ FLOP

Parameters

27.8M

Published

Oct 15, 2019

Rubik's cube ADR robot is an AI model developed by OpenAI (United States), first published in October 2019. It works in the robotics domain, on tasks such as robotic manipulation.

Training it took an estimated 8.5×10²⁰ FLOP of compute (estimation method: third-party estimation). The model has 27,769,565 parameters. Training ran on NVIDIA Tesla V100 DGXS 32 GB.

Access: Unreleased. Its weights are not openly released. The reference paper has 1,359 citations. Epoch AI rates the confidence of this record as likely.

Full record

Organization: OpenAI
Country of organization: United States
Domain: Robotics
Task: Robotic manipulation
Training compute: 8.5×10²⁰ FLOP
Compute estimation method: Third-party estimation
Parameters: 27,769,565
Training hardware: NVIDIA Tesla V100 DGXS 32 GB
Model accessibility: Unreleased
Open weights: No
Citations: 1,359
Epoch confidence: Likely