tensorflow model save and restore

TensorFlow 模型保存/载入

我们在上线使用一个算法模型的时候，首先必须将已经训练好的模型保存下来。tensorflow保存模型的方式与sklearn不太一样，sklearn很直接，一个sklearn.externals.joblib的dump与load方法就可以保存与载入使用。而tensorflow由于有graph, operation 这些概念，保存与载入模型稍显麻烦。由于TensorFlow 的版本一直在更新, 保存模型的方法也发生了改变，在python 环境,和在C++ 环境(移动端) 等不同的平台需要的模型文件也是不也一样的。

(https://stackoverflow.com/questions/44516609/tensorflow-what-is-the-relationship-between-ckpt-file-and-ckpt-meta-and-ckp有一个解释如下)

the .ckpt file is the old version output of saver.save(sess), which is the equivalent of your .ckpt-data (see below)
the "checkpoint" file is only here to tell some TF functions which is the latest checkpoint file.
.ckpt-meta contains the metagraph, i.e. the structure of your computation graph, without the values of the variables (basically what you can see in tensorboard/graph).
.ckpt-data contains the values for all the variables, without the structure. To restore a model in python, you'll usually use the meta and data files with (but you can also use the .pb file):

saver = tf.train.import_meta_graph(path_to_ckpt_meta)

saver.restore(sess, path_to_ckpt_data)

I don't know exactly for .ckpt-index, I guess it's some kind of index needed internally to map the two previous files correctly. Anyway it's not really necessary usually, you can restore a model with only .ckpt-meta and .ckpt-data.
the .pb file can save your whole graph (meta + data). To load and use (but not train) a graph in c++ you'll usually use it, created with freeze_graph, which creates the .pb file from the meta and data. Be careful, (at least in previous TF versions and for some people) the py function provided by freeze_graph did not work properly, so you'd have to use the script version. Tensorflow also provides a tf.train.Saver.to_proto() method, but I don't know what it does exactly.

一、基本方法

网上搜索tensorflow模型保存，搜到的大多是基本的方法。即

保存

定义变量
使用saver.save()方法保存

载入

定义变量
使用saver.restore()方法载入

如保存代码如下

import tensorflow as tf

import numpy as np  

W = tf.Variable([[1,1,1],[2,2,2]],dtype = tf.float32,name='w')

b = tf.Variable([[0,1,2]],dtype = tf.float32,name='b')  

init = tf.initialize_all_variables()

saver = tf.train.Saver()

with tf.Session() as sess:

        sess.run(init)

        save_path = saver.save(sess,"save/model.ckpt")

载入代码如下

import tensorflow as tf

import numpy as np  

W = tf.Variable(tf.truncated_normal(shape=(2,3)),dtype = tf.float32,name='w')

b = tf.Variable(tf.truncated_normal(shape=(1,3)),dtype = tf.float32,name='b')  

saver = tf.train.Saver()

with tf.Session() as sess:

        saver.restore(sess,"save/model.ckpt")

二、不需重新定义网络结构的方法

tf.train.import_meta_graph(

    meta_graph_or_file,

    clear_devices=False,

    import_scope=None,

    **kwargs

)

这个方法可以从文件中将保存的graph的所有节点加载到当前的default graph中，并返回一个saver。也就是说，我们在保存的时候，除了将变量的值保存下来，其实还有将对应graph中的各种节点保存下来，所以模型的结构也同样被保存下来了。

比如我们想要保存计算最后预测结果的y，则应该在训练阶段将它添加到collection中。具体代码如下

保存

### 定义模型

input_x = tf.placeholder(tf.float32, shape=(None, in_dim), name='input_x')

input_y = tf.placeholder(tf.float32, shape=(None, out_dim), name='input_y')

w1 = tf.Variable(tf.truncated_normal([in_dim, h1_dim], stddev=0.1), name='w1')

b1 = tf.Variable(tf.zeros([h1_dim]), name='b1')

w2 = tf.Variable(tf.zeros([h1_dim, out_dim]), name='w2')

b2 = tf.Variable(tf.zeros([out_dim]), name='b2')

keep_prob = tf.placeholder(tf.float32, name='keep_prob')

hidden1 = tf.nn.relu(tf.matmul(self.input_x, w1) + b1)

hidden1_drop = tf.nn.dropout(hidden1, self.keep_prob)

### 定义预测目标

y = tf.nn.softmax(tf.matmul(hidden1_drop, w2) + b2)

# 创建saver

saver = tf.train.Saver(...variables...)

# 假如需要保存y，以便在预测时使用

tf.add_to_collection('pred_network', y)

sess = tf.Session()

for step in xrange(1000000):

    sess.run(train_op)

    if step % 1000 == 0:

        # 保存checkpoint, 同时也默认导出一个meta_graph

        # graph名为'my-model-{global_step}.meta'.

        saver.save(sess, 'my-model', global_step=step)

载入

with tf.Session() as sess:

  new_saver = tf.train.import_meta_graph('my-save-dir/my-model-10000.meta')

  new_saver.restore(sess, 'my-save-dir/my-model-10000')

  # tf.get_collection() 返回一个list. 但是这里只要第一个参数即可

  y = tf.get_collection('pred_network')[0]

  graph = tf.get_default_graph()

  # 因为y中有placeholder，所以sess.run(y)的时候还需要用实际待预测的样本以及相应的参数来填充这些placeholder，而这些需要通过graph的get_operation_by_name方法来获取。

  input_x = graph.get_operation_by_name('input_x').outputs[0]

  keep_prob = graph.get_operation_by_name('keep_prob').outputs[0]

  # 使用y进行预测

  sess.run(y, feed_dict={input_x:....,  keep_prob:1.0})

这里有两点需要注意的：

一、 saver.restore()时填的文件名，因为在saver.save的时候，每个checkpoint会保存三个文件，如
my-model-10000.meta, my-model-10000.index, my-model-10000.data-00000-of-00001
在import_meta_graph时填的就是meta文件名，我们知道权值都保存在my-model-10000.data-00000-of-00001这个文件中，但是如果在restore方法中填这个文件名，就会报错，应该填的是前缀，这个前缀可以使用tf.train.latest_checkpoint(checkpoint_dir)这个方法获取。

二、模型的y中有用到placeholder，在sess.run()的时候肯定要feed对应的数据，因此还要根据具体placeholder的名字，从graph中使用get_operation_by_name方法获取。