CNTK Examples: Image/Classification/GoogLeNet/Inception-ResNet-V1

BrainScript

Inception-ResNet-V1.cntk

The Inception-ResNet-V1 model is implemented according to the model described in Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning.

This implementation achieves 78.210% Top-1 accuracy and 93.848% Top-5 accuracy in validation set.

And we're using common SGD with Nesterov momentum in this implementation.

This example with a 256 batch-size should be trained with 8 GPUs.

You could run the example from the current folder using:

mpiexec -n 8 cntk configFile=Inception-ResNet-V1.cntk

If you would like to run this example with a single card. You could divide the minibatchSize and learningRatesPerMB with a ratio 8 simultaneously.

And run this example using:

cntk configFile=Inception-ResNet-V1.cntk