Reading .caffemodel and mapping on DET

[sharath s]

Hello

I am pretty new caffe still, i successfully trained a dataset which i created using CifarNet example in Caffe.

After the training, framework has spit out files - cifar_quick_iter_5000.caffemodel and cifar10_quick_iter_5000.solverstate

i guess cifar_quick_iter_5000.caffemodel is the trained CNN Weights, How do i know how the trained network has performed.

I want to read .caffemodel and map it onto a Detection Error Tradeoff(DET) curve, how can i do this?

Thanks
BR
Sharath