使用pycaffe读取caffemodel参数（保存到txt文件）

import struct  
import numpy as np
import sys


def add_path(path):
    if path not in sys.path:
        sys.path.insert(0, path)
add_path('./python/')
import caffe 


np.set_printoptions(threshold='nan')


MODEL_FILE = 'test.prototxt'
PRETRAIN_FILE = 'caffe_iter_10000.caffemodel'


params_txt = 'params.txt'  
pf = open(params_txt, 'w') 


net = caffe.Net(MODEL_FILE, PRETRAIN_FILE, caffe.TEST) 
  
for param_name in net.params.keys():   
    weight = net.params[param_name][0].data   
    bias = net.params[param_name][1].data  
    
    pf.write(param_name)  
    pf.write('\n')  
   
    pf.write('\n' + param_name + '_weight:\n\n')   
    weight.shape = (-1, 1)  
  
    for w in weight:  
        pf.write('%ff, ' % w)  
    
    pf.write('\n\n' + param_name + '_bias:\n\n')   
    bias.shape = (-1, 1)  
    for b in bias:  
        pf.write('%ff, ' % b)  
  
    pf.write('\n\n')    
pf.close