segmentation fault(core dumped) when eval() about fold HOT 13 OPEN

I have it working with python 3.6 and tensorflow 1.0.0 (ubuntu 16.04)
1.0.1 caused seg-faults in eval for me too

from fold.

yes, i have the same problem with python 3.5

from fold.

even when building from source.

from fold.

from fold.

i'm on linux ; i'll try with tf-1.0.0 instead of 1.0.1

from fold.

i got it to run on linux with python 2.7 and tensorflow 1.0.0rc0

from fold.

moshelooks: Yes, with TF 1.0.0rc0, tensorflow-fold works now. Thank you!

from fold.

I have it working with python 2.7 and tensorflow 1.0.1 (ubuntu 16.04), it caused segfault too, on eval()

from fold.

I also have the same problem with Ubuntu 14.04 tensor flow 1.0.1 python 2.7

from fold.

I faced the same problem when using tensorflow 1.0.0. But after using 1.0.0-rc0-gpu installed with Docker, the segfault is gone. (now I have other issues, but at lease segfault is gone)
I am using Python 2.7, Ubuntu 14.04, TF 1.0.0-rc0

from fold.

Same on Ubuntu 16 LTS, Tensorflow-GPU 1.2.1 !

from fold.

Indeed, just spend hours trying to build appropriate image to get rid of segmentation fault and found this thread. I tried next combination:

scipy==0.19.1
h5py==2.7.1
numpy==1.13.1
tensorflow==1.0.0
tflearn==0.3.2

This works for sure, previous attempts to install all latest dependencies (especially tensorflow 1.3.0):

bleach==1.5.0
emorecognition==0.0.2
h5py==2.7.1
html5lib==0.9999999
Markdown==2.6.9
nose==1.3.7
numpy==1.13.1
olefile==0.44
pandas==0.20.3
Pillow==4.2.1
protobuf==3.4.0
python-dateutil==2.6.1
pytz==2017.2
scipy==0.19.1
six==1.10.0
tensorflow==1.3.0
tensorflow-tensorboard==0.1.6
tflearn==0.3.2
Werkzeug==0.12.2

Was causing a segfault 11. Now i'm able to build fine grained docker image using jjanzic/docker-python3-opencv as bases.

Unfortunately i have no clue what was going on before segfault killed docker container =/

from fold.

This issue still valid for tflearn tflearn/tflearn#905

from fold.