LeNet-5

Training compute

2.8×10¹² FLOP

Parameters

60K

Published

Nov 1, 1998

LeNet-5 is an AI model developed by AT&T (United States), first published in November 1998. It works in the vision domain, on tasks such as character recognition (ocr). It counts among the frontier models: the systems trained with the most compute of their moment.

Training it took an estimated 2.8×10¹² FLOP of compute (estimation method: operation counting). The model has 60,000 parameters. It was trained on roughly 60K datapoints.

The reference paper has 57,900 citations. Epoch AI rates the confidence of this record as confident.

Full record

Organization: AT&T
Country of organization: United States
Domain: Vision
Task: Character recognition (OCR)
Training compute: 2.8×10¹² FLOP
Compute estimation method: Operation counting
Parameters: 60,000
Dataset size: 60K
Citations: 57,900
Epoch confidence: Confident