Pooling CNN (NORB)

Training compute

1.5×10¹⁵ FLOP

Parameters

268.7K

Published

Sep 15, 2010

Pooling CNN (NORB) is an AI model developed by University of Bonn (Germany), first published in September 2010. It works in the vision domain, on tasks such as image classification.

Training it took an estimated 1.5×10¹⁵ FLOP of compute (estimation method: operation counting). The model has 268,664 parameters. It was trained on roughly 24.3K datapoints.

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

Full record

Organization: University of Bonn
Country of organization: Germany
Domain: Vision
Task: Image classification
Training compute: 1.5×10¹⁵ FLOP
Compute estimation method: Operation counting
Parameters: 268,664
Dataset size: 24.3K
Model accessibility: Unreleased
Open weights: No
Epoch confidence: Confident