CODEWITHSUNDEEP
pythonmachine-learningtensorflow
Luis Leal
Luis Leal

Reputation: 3514

how to store/save and restore tensorflow DNNClassifier(No variables to save)

I trained a deep neural network on tensorflow and used to predict some examples but, when I try to save it using train.Saver() I get the error: "No variables to save"

Already tried train.Saver like this:

tf.train.Saver(classi.get_variable_names())

But still no luck, any suggestions?

Upvotes: 1

Views: 2052

Answers (2)

craymichael
craymichael

Reputation: 4821

So I ran into the same issue (Estimators don't have save/restore functions yet). I tried savers and CheckpointSaver to try and save checkpoints but turns out it's much simpler; just specify the model_dir when instantiating the Estimator. This will automatically save checkpoints that can be restored simply by creating an Estimator with the same model_dir. Documentation for Estimators here.

Thanks to @ilblackdragon for the solution here.

Upvotes: 4

Shan Carter
Shan Carter

Reputation: 2455

Here's a code sample from the tf.variable docs that might clarify:

# Create some variables.
v1 = tf.Variable(..., name="v1")
v2 = tf.Variable(..., 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)
  # Do some work with the model.
  ..
  # Save the variables to disk.
  save_path = saver.save(sess, "/tmp/model.ckpt")
  print("Model saved in file: %s" % save_path)

Upvotes: -1

Related Questions