FaceBook ResNet

show more

Implementations

ResNet training in Torch

implements training of residual networks from Deep Residual Learning for Image Recognition by Kaiming He, et. al.

Requirements

See the installation instructions for a step-by-step guide.

If you already have Torch installed, update nn, cunn, and cudnn.

fbresnet152(num_classes=1000, pretrained='imagenet')

Pretrained

Trained ResNet 18, 34, 50, 101, 152, and 200 models are available for download.

 

Single-crop (224x224) validation error rate

NetworkTop-1 errorTop-5 error
ResNet-1830.4310.76
ResNet-3426.738.74
ResNet-5024.017.02
ResNet-10122.446.21
ResNet-15222.166.16
ResNet-20021.665.79
classificationtorch
How helpful was this page?