finetune_alexnet_with_tensorflow
finetune_alexnet_with_tensorflow copied to clipboard
A question about the structure
hi, I have a question. Why is that code in function create() in "https://kratzert.github.io/2017/02/24/finetuning-alexnet-with-tensorflow.html" the pool1 layer is after norm1 layer, and here norm1 is after pool1? As I know, the structure of alexnet in article is the former.
It is a long time since I wrote this code and I'm not sure if I remember correctly, but I think this is the caffenet version of AlexNet which switched the two layers.
thanks! I don't understand the difference of them. Maybe it just will get better results to switch them?