Wrong model definition

[Atcold]

googlenet.lua does not reflect GoogLeNet architecture (defined in Going Deeper with Convolutions).

In GoogLeNet, inception modules have 2 layers only with 3x3 and 5x5 convolution and 3x3 max-pooling, plus 1x1 dimension matching convolutions.

What is googlenet.lua implementing?

I think there is a little of confusion with the models. From Szegedy we have:

• GoogLeNet :x:
• BN-GoogLeNet
• BN-Inception
• Inception-v2
  • Vanilla
  • Label Smoothing
  • Factorized 7 × 7
  • BN-auxiliary
• Inception-v3

and GoogLeNet is not the implemented architecture.

[Atcold]

Here there is a better (more correct) version, but still it is not GoogLeNet.

[Atcold]

Here there is a reliable implementation -> https://github.com/e-lab/imagenet-multiGPU.torch/blob/trueGoogLeNet/models/googlenetFixed_cudnn.lua