NVAE (Celeba HQ)

Training compute

3×10²⁰ FLOP

Published

Jan 8, 2021

NVAE (Celeba HQ) is an AI model developed by NVIDIA (United States), first published in January 2021. It works in the image generation domain, on tasks such as image generation.

Training it took an estimated 3×10²⁰ FLOP of compute (estimation method: hardware). Training ran on 24 NVIDIA Tesla V100 DGXS 32 GB for about 94 hours.

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

Full record

Organization: NVIDIA
Country of organization: United States
Domain: Image generation
Task: Image generation
Training compute: 3×10²⁰ FLOP
Compute estimation method: Hardware
Training hardware: NVIDIA Tesla V100 DGXS 32 GB
Chips used: 24
Training time: 94 h
Training power draw: 12.2 kW
Model accessibility: Unreleased
Open weights: No
Epoch confidence: Confident