Tensorboard Visual Debugger UnimplementedError object not iterable

Question

My model is able to run fine until I try and use the visual debugger. When I remove tf_debug.TensorBoardDebugWrapperSession the model runs fine.

Traceback (most recent call last): File "...", line 337, in feed_dict={X: X_mb, z: sample_z(mb_size, z_dim)} TypeError: 'UnimplementedError' object is not iterable

If more context is needed for my code I will accordingly edit the following.

import tensorflow as tf
from tensorflow.python import debug as tf_debug
sess = tf.Session(config=config)
sess = tf_debug.TensorBoardDebugWrapperSession(sess, "MYPORT")
...
...
_, D_loss_curr, _ = sess.run(
            [D_solver, D_loss, clip_D],
                feed_dict={X: X_mb, z: sample_z(mb_size, z_dim)}
        )

Answer 1

It seems to be a problem on windows, see open issue: https://github.com/tensorflow/tensorflow/issues/17933