RetinaNet-R101

Training compute

2.1×10¹⁸ FLOP

Parameters

53M

Published

Aug 7, 2017

RetinaNet-R101 is an AI model developed by Facebook AI Research (United States and France), first published in August 2017. It works in the vision domain, on tasks such as object detection.

Training it took an estimated 2.1×10¹⁸ FLOP of compute (estimation method: hardware). The model has 53,000,000 parameters. It was trained on roughly 115K datapoints. Training ran on 8 NVIDIA M40 for about 35 hours.

The reference paper has 16,437 citations. Epoch AI rates the confidence of this record as confident.

Full record

Organization: Facebook AI Research
Country of organization: United States, France
Domain: Vision
Task: Object detection
Training compute: 2.1×10¹⁸ FLOP
Compute estimation method: Hardware
Parameters: 53,000,000
Dataset size: 115K
Training hardware: NVIDIA M40
Chips used: 8
Training time: 35 h
Chip-hours: 280
Training power draw: 4.2 kW
Numerical format: FP32
Citations: 16,437
Epoch confidence: Confident