DALL-E mega

Training compute

2.3×10²² FLOP

Published

Jun 28, 2022

DALL-E mega is an AI model developed by Craiyon (United States), first published in June 2022. It works in the image generation domain, on tasks such as text-to-image and image generation.

Training it took an estimated 2.3×10²² FLOP of compute (estimation method: hardware). Training ran on 128 Google TPU v3 for about 1.3K hours.

Access: Open weights (unrestricted). Its weights are openly available. Epoch AI rates the confidence of this record as confident.

Full record

Organization: Craiyon
Country of organization: United States
Domain: Image generation
Task: Text-to-image, Image generation
Training compute: 2.3×10²² FLOP
Compute estimation method: Hardware
Training hardware: Google TPU v3
Chips used: 128
Training time: 1,344 h
Chip-hours: 172K
Training power draw: 115.5 kW
Model accessibility: Open weights (unrestricted)
Open weights: Yes
Epoch confidence: Confident