Caffe training/testing

• Training LeNet with MNIST dataset
• http://caffe.berkeleyvision.org/gathered/examples/mnist.html
• this is a simpler model with small dataset to train it quickly even on a CPU
• change the 'solver_mode' value  in the proto buffer file  lenet_solver.prototxt in order to train on CPU only machine                 # solver mode: CPU or GPU
  solver_mode: CPU
• 
  
•  Training XNet on ImageNet dataset
• download existing trained .caffemodel
• ./script/download_model_binary.py model/blvc_reference_rcnn_ilsvrc13
• should download blvc_reference_rcnn_ilsvrc13.caffemodel
• it looks for the following two files "deploy.prototxt, readme.md" inside the models/bvlc_reference_rcnn_ilsvrc13 directory
• training new caffe model using own data 
•  
•  
•