MuZero

Training compute

4.8×10¹⁹ FLOP

Parameters

36.9M

Published

Nov 19, 2019

MuZero is an AI model developed by DeepMind (United Kingdom), first published in November 2019. It works in the games domain, on tasks such as atari.

Training it took an estimated 4.8×10¹⁹ FLOP of compute (estimation method: hardware). The model has 36,864,000 parameters. It was trained on roughly 12.3B datapoints.

Access: Unreleased. Its weights are not openly released. The reference paper has 2,491 citations. Epoch AI rates the confidence of this record as speculative.

Full record

Organization: DeepMind
Country of organization: United Kingdom
Domain: Games
Task: Atari
Training compute: 4.8×10¹⁹ FLOP
Compute estimation method: Hardware
Parameters: 36,864,000
Dataset size: 12.3B
Model accessibility: Unreleased
Open weights: No
Citations: 2,491
Epoch confidence: Speculative