This model is a re-implementation of Batch Normalization publication, and the model is trained with a customized caffe; however, the modifications are minor. Thus, you can run this with the currently available official caffe version, including cudnn v4 support and multigpu support.
The uploaded caffemodel is the snapshot of 1,200,000 iteration (30 epochs) using solver_stepsize_6400.prototxt
The uploaded model achieves a top-1 accuracy 72.05% (27.95% error) and a top-5 accuracy 90.87% (9.13% error) on the validation set, using a single center crop.
Thank John Lee for helping me training this model.
Tips for performance
Real-time data shuffling is important
Data augmentation during training should improve the accuracy.
Change interpolation method (default is bilinear) of opencv to bicubic when you convert image will give you minor improvement.