GAIA-1

Training compute

1.2×10²² FLOP

Parameters

9B

Published

Sep 29, 2023

GAIA-1 is an AI model developed by Wayve (United Kingdom), first published in September 2023. It works in the video, vision, multimodal and language domain, on tasks such as self-driving car, video generation and instruction interpretation.

Training it took an estimated 1.2×10²² FLOP of compute (estimation method: hardware,reported). The model has 9,000,000,000 parameters. Training ran on NVIDIA A100.

Access: Unreleased. Its weights are not openly released. Epoch AI rates the confidence of this record as confident.

Full record

Organization: Wayve
Country of organization: United Kingdom
Domain: Video, Vision, Multimodal, Language
Task: Self-driving car, Video generation, Instruction interpretation
Training compute: 1.2×10²² FLOP
Compute estimation method: Hardware, Reported
Parameters: 9,000,000,000
Training hardware: NVIDIA A100
Chip-hours: 34.6K
Model accessibility: Unreleased
Open weights: No
Epoch confidence: Confident