DALL-E mini

Training compute

3.8×10¹⁹ FLOP

Published

Oct 26, 2021

DALL-E mini is an AI model developed by Craiyon (United States), first published in October 2021. It works in the vision domain, on tasks such as text-to-image.

Training it took an estimated 3.8×10¹⁹ FLOP of compute (estimation method: hardware). It was trained on roughly 4.4B datapoints. Training ran on 4 Google TPU v3 for about 72 hours.

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

Full record

Organization: Craiyon
Country of organization: United States
Domain: Vision
Task: Text-to-image
Training compute: 3.8×10¹⁹ FLOP
Compute estimation method: Hardware
Dataset size: 4.4B
Training hardware: Google TPU v3
Chips used: 4
Training time: 72 h
Chip-hours: 288
Training power draw: 3.6 kW
Model accessibility: Open weights (unrestricted)
Open weights: Yes
Epoch confidence: Likely