class saver

作用：用于保存和恢复网络的参数

Function

Init

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

__init__(
    var_list=None,
    reshape=False,
    sharded=False,
    max_to_keep=5,
    keep_checkpoint_every_n_hours=10000.0,
    name=None,
    restore_sequentially=False,
    saver_def=None,
    builder=None,
    defer_build=False,
    allow_empty=False,
    write_version=tf.train.SaverDef.V2,
    pad_step_number=False,
    save_relative_paths=False,
    filename=None
)

Args:

var_list: 指定保存的变量。可以使list或者dict

max_to_keep: 最近可以保存ckeckpoints的最大数目

keep_checkpoint_every_n_hours： 多久保存一次ckeckpoints

1
2
3
4
5
6
7
8
9
10
11
12
13


v1 = tf.Variable(..., name='v1')
v2 = tf.Variable(..., name='v2')

# Pass the variables as a dict:
saver = tf.train.Saver({'v1': v1, 'v2': v2})

# Or pass them as a list.
saver = tf.train.Saver([v1, v2])
# Passing a list is equivalent to passing a dict with the variable op names
# as keys:
saver = tf.train.Saver({v.op.name: v for v in [v1, v2]})

save

1
2
3
4
5
6
7
8
9
10
save(
    sess,
    save_path,
    global_step=None,
    latest_filename=None,
    meta_graph_suffix='meta',
    write_meta_graph=True,
    write_state=True,
    strip_default_attrs=False
)

作用：保存网络参数

Args:
sess: 会话句柄
save_path：保存路径
global_step：步数

restore

作用：恢复网络参数

1
2
3
4
restore(
    sess,
    save_path
)

Args:
sess: 会话句柄
save_path：保存路径

Example

1
2
3
4
5
6
7
8
9
# Create a saver.
saver = tf.train.Saver(...variables...)
# Launch the graph and train, saving the model every 1,000 steps.
sess = tf.Session()
for step in xrange(1000000):
    sess.run(..training_op..)
    if step % 1000 == 0:
        # Append the step number to the checkpoint name:
        saver.save(sess, 'my-model', global_step=step)

查看保存文件里的tensor

1
2
3
4
5
6
7
8
9
from tensorflow.python import pywrap_tensorflow

# 从检查点文件中读取数据
reader = pywrap_tensorflow.NewCheckpointReader(checkpoint_path)
var_to_shape_map = reader.get_variable_to_shape_map()
# 显示变量名及其值
for key in var_to_shape_map:
    print("tensor_name: ", key)
    print(reader.get_tensor(key))

保存文件：

  • .meta文件保存了当前图结构
  • .index文件保存了当前参数名
  • .data文件保存了当前参数值

注：每调用一次save方法会产生新的文件

To do

  • transfer learning 里如何恢复部分模型参数