python----------->>>>>>>>>>>>统计caffe模型的模型参数量

?

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

#统计caffe算模型的参数量 from numpy import prod, sum flops = 0 typenames = ['Convolution', 'BatchNorm'] for layer_name, blob in net.blobs.iteritems():     if layer_name not in net.layer_dict:         continue     if net.layer_dict[layer_name].type in typenames:         cur_flops = 0.0         if net.layer_dict[layer_name].type in typenames[:2]:             cur_flops = (np.product(net.params[layer_name][0].data.shape) * \                     blob.data.shape[-1] * blob.data.shape[-2])         else:             cur_flops = np.product(net.params[layer_name][0].data.shape)         print(layer_name.ljust(20),                 str(net.params[layer_name][0].data.shape).ljust(20),                 str(blob.data.shape).ljust(20),                 net.layer_dict[layer_name].type.ljust(20), str(cur_flops).ljust(20))         # InnerProduct         if len(blob.data.shape) == 2:             flops += prod(net.params[layer_name][0].data.shape)         else:             flops += prod(net.params[layer_name][0].data.shape) * blob.data.shape[2] * blob.data.shape[3]   print ('layers num: ' + str(len(net.params.items()))) print ("Total number of parameters: " + str(sum([prod(v[0].data.shape) for k, v in net.params.items()]))) print ("Total number of flops: " + str(flops))