恢复TensorFlow模型

Question

I'm trying to restore TensorFlow model. I followed this example: http://nasdag.github.io/blog/2016/01/19/classifying-bees-with-google-tensorflow/

At the end of the code in the example I added these lines:

saver = tf.train.Saver()
save_path = saver.save(sess, "model.ckpt")
print("Model saved in file: %s" % save_path)

Two files were created: checkpoint and model.ckpt.

In a new python file (tomas_bees_predict.py), I have this code:

import tensorflow as tf

saver = tf.train.Saver()

with tf.Session() as sess:
  # Restore variables from disk.
  saver.restore(sess, "model.ckpt")
  print("Model restored.")

However when I execute the code, I get this error:

Traceback (most recent call last):
  File "tomas_bees_predict.py", line 3, in <module>
    saver = tf.train.Saver()
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/training/saver.py", line 705, in __init__
raise ValueError("No variables to save")

ValueError: No variables to save

Is there a way to read mode.ckpt file and see what variables are saved? Or maybe someone can help with saving the model and restoring it based on the example described above?

EDIT 1:

I think I tried running the same code in order to recreate model structure and I was getting the error. I think it could be related to the fact that code described here isn't using named variables: http://nasdag.github.io/blog/2016/01/19/classifying-bees-with-google-tensorflow/

def weight_variable(shape):
  initial = tf.truncated_normal(shape, stddev=0.1)
  return tf.Variable(initial)

def bias_variable(shape):
  initial = tf.constant(0.1, shape=shape)
  return tf.Variable(initial)

So I did this experiment. I wrote two versions of the code (with and without named variables) to save the model and the code to restore the model.

tensor_save_named_vars.py :

import tensorflow as tf

# Create some variables.
v1 = tf.Variable(1, name="v1")
v2 = tf.Variable(2, name="v2")

# Add an op to initialize the variables.
init_op = tf.initialize_all_variables()

# Add ops to save and restore all the variables.
saver = tf.train.Saver()

# Later, launch the model, initialize the variables, do some work, save the
# variables to disk.
with tf.Session() as sess:
  sess.run(init_op)
  print "v1 = ", v1.eval()
  print "v2 = ", v2.eval()
  # Save the variables to disk.
  save_path = saver.save(sess, "/tmp/model.ckpt")
  print "Model saved in file: ", save_path

tensor_save_not_named_vars.py:

import tensorflow as tf

# Create some variables.
v1 = tf.Variable(1)
v2 = tf.Variable(2)

# Add an op to initialize the variables.
init_op = tf.initialize_all_variables()

# Add ops to save and restore all the variables.
saver = tf.train.Saver()

# Later, launch the model, initialize the variables, do some work, save the
# variables to disk.
with tf.Session() as sess:
  sess.run(init_op)
  print "v1 = ", v1.eval()
  print "v2 = ", v2.eval()
  # Save the variables to disk.
  save_path = saver.save(sess, "/tmp/model.ckpt")
  print "Model saved in file: ", save_path

tensor_restore.py:

import tensorflow as tf

# Create some variables.
v1 = tf.Variable(0, name="v1")
v2 = tf.Variable(0, name="v2")

# Add ops to save and restore all the variables.
saver = tf.train.Saver()

# Later, launch the model, use the saver to restore variables from disk, and
# do some work with the model.
with tf.Session() as sess:
  # Restore variables from disk.
  saver.restore(sess, "/tmp/model.ckpt")
  print "Model restored."
  print "v1 = ", v1.eval()
  print "v2 = ", v2.eval()

Here is what I get when I execute this code:

$ python tensor_save_named_vars.py 

I tensorflow/core/common_runtime/local_device.cc:40] Local device intra op parallelism threads: 4
I tensorflow/core/common_runtime/direct_session.cc:58] Direct session inter op parallelism threads: 4
v1 =  1
v2 =  2
Model saved in file:  /tmp/model.ckpt

$ python tensor_restore.py 

I tensorflow/core/common_runtime/local_device.cc:40] Local device intra op parallelism threads: 4
I tensorflow/core/common_runtime/direct_session.cc:58] Direct session inter op parallelism threads: 4
Model restored.
v1 =  1
v2 =  2

$ python tensor_save_not_named_vars.py 

I tensorflow/core/common_runtime/local_device.cc:40] Local device intra op parallelism threads: 4
I tensorflow/core/common_runtime/direct_session.cc:58] Direct session inter op parallelism threads: 4
v1 =  1
v2 =  2
Model saved in file:  /tmp/model.ckpt

$ python tensor_restore.py 
I tensorflow/core/common_runtime/local_device.cc:40] Local device intra op parallelism threads: 4
I tensorflow/core/common_runtime/direct_session.cc:58] Direct session inter op parallelism threads: 4
W tensorflow/core/common_runtime/executor.cc:1076] 0x7ff953881e40 Compute status: Not found: Tensor name "v2" not found in checkpoint files /tmp/model.ckpt
     [[Node: save/restore_slice_1 = RestoreSlice[dt=DT_INT32, preferred_shard=-1, _device="/job:localhost/replica:0/task:0/cpu:0"](_recv_save/Const_0, save/restore_slice_1/tensor_name, save/restore_slice_1/shape_and_slice)]]
W tensorflow/core/common_runtime/executor.cc:1076] 0x7ff953881e40 Compute status: Not found: Tensor name "v1" not found in checkpoint files /tmp/model.ckpt
     [[Node: save/restore_slice = RestoreSlice[dt=DT_INT32, preferred_shard=-1, _device="/job:localhost/replica:0/task:0/cpu:0"](_recv_save/Const_0, save/restore_slice/tensor_name, save/restore_slice/shape_and_slice)]]
Traceback (most recent call last):
  File "tensor_restore.py", line 14, in <module>
    saver.restore(sess, "/tmp/model.ckpt")
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/training/saver.py", line 891, in restore
    sess.run([self._restore_op_name], {self._filename_tensor_name: save_path})
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 368, in run
    results = self._do_run(target_list, unique_fetch_targets, feed_dict_string)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 444, in _do_run
    e.code)
tensorflow.python.framework.errors.NotFoundError: Tensor name "v2" not found in checkpoint files /tmp/model.ckpt
     [[Node: save/restore_slice_1 = RestoreSlice[dt=DT_INT32, preferred_shard=-1, _device="/job:localhost/replica:0/task:0/cpu:0"](_recv_save/Const_0, save/restore_slice_1/tensor_name, save/restore_slice_1/shape_and_slice)]]
Caused by op u'save/restore_slice_1', defined at:
  File "tensor_restore.py", line 8, in <module>
    saver = tf.train.Saver()
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/training/saver.py", line 713, in __init__
    restore_sequentially=restore_sequentially)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/training/saver.py", line 432, in build
    filename_tensor, vars_to_save, restore_sequentially, reshape)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/training/saver.py", line 191, in _AddRestoreOps
    values = self.restore_op(filename_tensor, vs, preferred_shard)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/training/saver.py", line 106, in restore_op
    preferred_shard=preferred_shard)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/ops/io_ops.py", line 189, in _restore_slice
    preferred_shard, name=name)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/ops/gen_io_ops.py", line 271, in _restore_slice
    preferred_shard=preferred_shard, name=name)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/ops/op_def_library.py", line 664, in apply_op
    op_def=op_def)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 1834, in create_op
    original_op=self._default_original_op, op_def=op_def)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 1043, in __init__
    self._traceback = _extract_stack()

So perhaps the original code (see the external link above) could be modified to something like this:

def weight_variable(shape):
  initial = tf.truncated_normal(shape, stddev=0.1)
  weight_var = tf.Variable(initial, name="weight_var")
  return weight_var

def bias_variable(shape):
  initial = tf.constant(0.1, shape=shape)
  bias_var = tf.Variable(initial, name="bias_var")
  return bias_var

But then the question I have: is restoring weight_var and bias_var variables sufficient to implement the prediction? I did the training on the powerful machine with GPU and I would like to copy the model to the less powerful computer without GPU to run predictions.

Answer 1

There's a similar question here: Tensorflow: how to save/restore a model? TLDR; you need to recreate model structure using same sequence of TensorFlow API commands before using Saver object to restore the weights

This is suboptimal, follow Github issue #696 for progress on making this easier

Answer 2

This issue should be caused by the name scope variants when double creating the same network.

put the command:

tf.reset_default_graph()

before creating the network

Answer 3

If a problem like this occurs then try to restart your kernel as the current variable overwrites the previous causing conflict between them, thus it shows notFoundError and other issues come up.

I encountered the same type of problem and restarting the kernel worked for me. (Caution: Try avoiding running your kernel multiple times as it can ruin your model file recreating variables that overwrite the existing one thus end up changing the original values.)

Answer 4

I think I tried running the same code in order to recreate model structure and I was getting the error. I think it could be related to the fact that code described here isn't using named variables: http://nasdag.github.io/blog/2016/01/19/classifying-bees-with-google-tensorflow/

def weight_variable(shape):
  initial = tf.truncated_normal(shape, stddev=0.1)
  return tf.Variable(initial)

def bias_variable(shape):
  initial = tf.constant(0.1, shape=shape)
  return tf.Variable(initial)

So I did this experiment. I wrote two versions of the code (with and without named variables) to save the model and the code to restore the model.

tensor_save_named_vars.py :

import tensorflow as tf

# Create some variables.
v1 = tf.Variable(1, name="v1")
v2 = tf.Variable(2, name="v2")

# Add an op to initialize the variables.
init_op = tf.initialize_all_variables()

# Add ops to save and restore all the variables.
saver = tf.train.Saver()

# Later, launch the model, initialize the variables, do some work, save the
# variables to disk.
with tf.Session() as sess:
  sess.run(init_op)
  print "v1 = ", v1.eval()
  print "v2 = ", v2.eval()
  # Save the variables to disk.
  save_path = saver.save(sess, "/tmp/model.ckpt")
  print "Model saved in file: ", save_path

tensor_save_not_named_vars.py:

import tensorflow as tf

# Create some variables.
v1 = tf.Variable(1)
v2 = tf.Variable(2)

# Add an op to initialize the variables.
init_op = tf.initialize_all_variables()

# Add ops to save and restore all the variables.
saver = tf.train.Saver()

# Later, launch the model, initialize the variables, do some work, save the
# variables to disk.
with tf.Session() as sess:
  sess.run(init_op)
  print "v1 = ", v1.eval()
  print "v2 = ", v2.eval()
  # Save the variables to disk.
  save_path = saver.save(sess, "/tmp/model.ckpt")
  print "Model saved in file: ", save_path

tensor_restore.py:

import tensorflow as tf

# Create some variables.
v1 = tf.Variable(0, name="v1")
v2 = tf.Variable(0, name="v2")

# Add ops to save and restore all the variables.
saver = tf.train.Saver()

# Later, launch the model, use the saver to restore variables from disk, and
# do some work with the model.
with tf.Session() as sess:
  # Restore variables from disk.
  saver.restore(sess, "/tmp/model.ckpt")
  print "Model restored."
  print "v1 = ", v1.eval()
  print "v2 = ", v2.eval()

Here is what I get when I execute this code:

$ python tensor_save_named_vars.py 

I tensorflow/core/common_runtime/local_device.cc:40] Local device intra op parallelism threads: 4
I tensorflow/core/common_runtime/direct_session.cc:58] Direct session inter op parallelism threads: 4
v1 =  1
v2 =  2
Model saved in file:  /tmp/model.ckpt

$ python tensor_restore.py 

I tensorflow/core/common_runtime/local_device.cc:40] Local device intra op parallelism threads: 4
I tensorflow/core/common_runtime/direct_session.cc:58] Direct session inter op parallelism threads: 4
Model restored.
v1 =  1
v2 =  2

$ python tensor_save_not_named_vars.py 

I tensorflow/core/common_runtime/local_device.cc:40] Local device intra op parallelism threads: 4
I tensorflow/core/common_runtime/direct_session.cc:58] Direct session inter op parallelism threads: 4
v1 =  1
v2 =  2
Model saved in file:  /tmp/model.ckpt

$ python tensor_restore.py 
I tensorflow/core/common_runtime/local_device.cc:40] Local device intra op parallelism threads: 4
I tensorflow/core/common_runtime/direct_session.cc:58] Direct session inter op parallelism threads: 4
W tensorflow/core/common_runtime/executor.cc:1076] 0x7ff953881e40 Compute status: Not found: Tensor name "v2" not found in checkpoint files /tmp/model.ckpt
     [[Node: save/restore_slice_1 = RestoreSlice[dt=DT_INT32, preferred_shard=-1, _device="/job:localhost/replica:0/task:0/cpu:0"](_recv_save/Const_0, save/restore_slice_1/tensor_name, save/restore_slice_1/shape_and_slice)]]
W tensorflow/core/common_runtime/executor.cc:1076] 0x7ff953881e40 Compute status: Not found: Tensor name "v1" not found in checkpoint files /tmp/model.ckpt
     [[Node: save/restore_slice = RestoreSlice[dt=DT_INT32, preferred_shard=-1, _device="/job:localhost/replica:0/task:0/cpu:0"](_recv_save/Const_0, save/restore_slice/tensor_name, save/restore_slice/shape_and_slice)]]
Traceback (most recent call last):
  File "tensor_restore.py", line 14, in <module>
    saver.restore(sess, "/tmp/model.ckpt")
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/training/saver.py", line 891, in restore
    sess.run([self._restore_op_name], {self._filename_tensor_name: save_path})
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 368, in run
    results = self._do_run(target_list, unique_fetch_targets, feed_dict_string)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 444, in _do_run
    e.code)
tensorflow.python.framework.errors.NotFoundError: Tensor name "v2" not found in checkpoint files /tmp/model.ckpt
     [[Node: save/restore_slice_1 = RestoreSlice[dt=DT_INT32, preferred_shard=-1, _device="/job:localhost/replica:0/task:0/cpu:0"](_recv_save/Const_0, save/restore_slice_1/tensor_name, save/restore_slice_1/shape_and_slice)]]
Caused by op u'save/restore_slice_1', defined at:
  File "tensor_restore.py", line 8, in <module>
    saver = tf.train.Saver()
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/training/saver.py", line 713, in __init__
    restore_sequentially=restore_sequentially)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/training/saver.py", line 432, in build
    filename_tensor, vars_to_save, restore_sequentially, reshape)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/training/saver.py", line 191, in _AddRestoreOps
    values = self.restore_op(filename_tensor, vs, preferred_shard)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/training/saver.py", line 106, in restore_op
    preferred_shard=preferred_shard)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/ops/io_ops.py", line 189, in _restore_slice
    preferred_shard, name=name)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/ops/gen_io_ops.py", line 271, in _restore_slice
    preferred_shard=preferred_shard, name=name)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/ops/op_def_library.py", line 664, in apply_op
    op_def=op_def)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 1834, in create_op
    original_op=self._default_original_op, op_def=op_def)
  File "/usr/local/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 1043, in __init__
    self._traceback = _extract_stack()

So perhaps the original code (see the external link above) could be modified to something like this:

def weight_variable(shape):
  initial = tf.truncated_normal(shape, stddev=0.1)
  weight_var = tf.Variable(initial, name="weight_var")
  return weight_var

def bias_variable(shape):
  initial = tf.constant(0.1, shape=shape)
  bias_var = tf.Variable(initial, name="bias_var")
  return bias_var

But then the question I have: is restoring weight_var and bias_var variables sufficient to implement the prediction? I did the training on the powerful machine with GPU and I would like to copy the model to the less powerful computer without GPU to run predictions.

Answer 5

确保tf.train.Saver（）的声明与tf.Session（）一起使用