import tensorflow as tf
import numpy as np
import os
import h5py
import pickle
from tensorflow.python.framework import graph_util
from tensorflow.python.platform import gfile
#设置使用指定GPU
os.environ['CUDA_VISIBLE_DEVICES'] = '1'
#下面这段代码是在训练好之后将所有的权重名字和权重值罗列出来，训练的时候需要注释掉
reader = tf.train.NewCheckpointReader('./model.ckpt-100')
variables = reader.get_variable_to_shape_map()
for ele in variables:
  print(ele)
  print(reader.get_tensor(ele))


x = tf.placeholder(tf.float32, shape=[None, 1])
y = 4 * x + 4

w = tf.Variable(tf.random_normal([1], -1, 1))
b = tf.Variable(tf.zeros([1]))
y_predict = w * x + b


loss = tf.reduce_mean(tf.square(y - y_predict))
optimizer = tf.train.GradientDescentOptimizer(0.5)
train = optimizer.minimize(loss)

isTrain = False#设成True去训练模型
train_steps = 100
checkpoint_steps = 50
checkpoint_dir = ''


saver = tf.train.Saver() # defaults to saving all variables - in this case w and b
x_data = np.reshape(np.random.rand(10).astype(np.float32), (10, 1))

with tf.Session() as sess:
  sess.run(tf.global_variables_initializer())
  if isTrain:
    for i in xrange(train_steps):
      sess.run(train, feed_dict={x: x_data})
      if (i + 1) % checkpoint_steps == 0:
        saver.save(sess, checkpoint_dir + 'model.ckpt', global_step=i+1)
  else:
    ckpt = tf.train.get_checkpoint_state(checkpoint_dir)
    if ckpt and ckpt.model_checkpoint_path:
      saver.restore(sess, ckpt.model_checkpoint_path)
    else:
      pass   
    print(sess.run(w))
    print(sess.run(b))
    graph_def = tf.get_default_graph().as_graph_def()
    #通过修改下面的函数，个人觉得理论上能够实现修改权重，但是很复杂，如果哪位有好办法，欢迎指教
    output_graph_def = graph_util.convert_variables_to_constants(sess, graph_def, ['Variable'])
    with tf.gfile.FastGFile('./test.pb', 'wb') as f:
      f.write(output_graph_def.SerializeToString())


with tf.Session() as sess:
#对应最后一部分的写，这里能够将对应的变量取出来
  with gfile.FastGFile('./test.pb', 'rb') as f:
    graph_def = tf.GraphDef()
    graph_def.ParseFromString(f.read())
  res = tf.import_graph_def(graph_def, return_elements=['Variable:0'])
  print(sess.run(res))
  print(sess.run(graph_def))