Training results in NaN (commit 2b03725)

[bliep]

Running train.py --nr_gpu=1 results in NaN for the loss from bits_per_dim on tensorflow 1.5.0.
With some API changes (keepdims -> keep_dims, axis= -> ) the same NaN occurs on tensorflow 1.4.x. Disabling the data driven initialization of the weight normalization (commenting out sess.run(init_pass, feed_dict)) however does not result in NaN loss.
Any clues?

[yang-song]

I have the same problem.

[learning-luke]

I'm also getting an error like this. What did you change axis= into, by the way? I changed it to dim= to get it to work on tf 1.4.

[attaullah]

same issue here.