TypeError: Fetch argument has invalid type float32, must be a string or Tensor

Question

I'm training a CNN quite similar to the one in this example, for image segmentation. The images are 1500x1500x1, and labels are of the same size.

After defining the CNN structure, and in launching the session as in this code sample: (conv_net_test.py)

with tf.Session() as sess:
sess.run(init)
summ = tf.train.SummaryWriter('/tmp/logdir/', sess.graph_def)
step = 1
print ("import data, read from read_data_sets()...")

#Data defined by me, returns a DataSet object with testing and training images and labels for segmentation problem.
data = import_data_test.read_data_sets('Dataset')

# Keep training until reach max iterations
while step * batch_size < training_iters:
    batch_x, batch_y = data.train.next_batch(batch_size)
    print ("running backprop for step %d" % step)
    batch_x = batch_x.reshape(batch_size, n_input, n_input, n_channels)
    batch_y = batch_y.reshape(batch_size, n_input, n_input, n_channels)
    batch_y = np.int64(batch_y)
    sess.run(optimizer, feed_dict={x: batch_x, y: batch_y, keep_prob: dropout})
    if step % display_step == 0:
        # Calculate batch loss and accuracy
        #pdb.set_trace()
        loss, acc = sess.run([loss, accuracy], feed_dict={x: batch_x, y: batch_y, keep_prob: 1.})
    step += 1
print "Optimization Finished"

I hit upon the following TypeError (stacktrace below):

    conv_net_test.py in <module>()
    178             #pdb.set_trace()
--> 179             loss, acc = sess.run([loss, accuracy], feed_dict={x: batch_x, y: batch_y, keep_prob: 1.})
    180         step += 1
    181     print "Optimization Finished!"

tensorflow/python/client/session.pyc in run(self, fetches, feed_dict, options, run_metadata)
    370     try:
    371       result = self._run(None, fetches, feed_dict, options_ptr,
--> 372                          run_metadata_ptr)
    373       if run_metadata:
    374         proto_data = tf_session.TF_GetBuffer(run_metadata_ptr)

tensorflow/python/client/session.pyc in _run(self, handle, fetches, feed_dict, options, run_metadata)
    582 
    583     # Validate and process fetches.
--> 584     processed_fetches = self._process_fetches(fetches)
    585     unique_fetches = processed_fetches[0]
    586     target_list = processed_fetches[1]

tensorflow/python/client/session.pyc in _process_fetches(self, fetches)
    538           raise TypeError('Fetch argument %r of %r has invalid type %r, '
    539                           'must be a string or Tensor. (%s)'
--> 540                           % (subfetch, fetch, type(subfetch), str(e)))

TypeError: Fetch argument 1.4415792e+2 of 1.4415792e+2 has invalid type <type 'numpy.float32'>, must be a string or Tensor. (Can not convert a float32 into a Tensor or Operation.)

I am stumped at this point. Maybe this is a simple case of converting the type, but I'm not sure how/where. Also, why does the loss have to be a string? (Assuming the same error will pop up for the accuracy as well, once this is fixed).

Any help appreciated!

Answer 1

Where you use loss = sess.run(loss), you redefine in python the variable loss.

The first time it will run fine. The second time, you will try to do:

sess.run(1.4415792e+2)

Because loss is now a float.

---

You should use different names like:

loss_val, acc = sess.run([loss, accuracy], feed_dict={x: batch_x, y: batch_y, keep_prob: 1.})