ResNet-200

Training compute

3×10¹⁹ FLOP

Published

Sep 17, 2016

ResNet-200 is an AI model developed by Microsoft Research Asia (China), first published in September 2016. It works in the vision domain, on tasks such as image classification.

Training it took an estimated 3×10¹⁹ FLOP of compute (estimation method: hardware). It was trained on roughly 1.3M datapoints.

Access: Unreleased. Its weights are not openly released. The reference paper has 10,822 citations. Epoch AI rates the confidence of this record as speculative.

Full record

Organization: Microsoft Research Asia
Country of organization: China
Domain: Vision
Task: Image classification
Training compute: 3×10¹⁹ FLOP
Compute estimation method: Hardware
Dataset size: 1.3M
Training time: 500 h
Model accessibility: Unreleased
Open weights: No
Citations: 10,822
Epoch confidence: Speculative