bpnet/deepexplain issue

[spour]

Hi,
Hope you're doing well! I was recently reading the bpnet paper (excellent work!), and was really interested in trying to use the contribution scores of deepexplain's implementation of deeplift with tf-modisco to create a motif. I posted this problem on the deepexplain page as well, but in case you have some insights, I received this error from the example code for the MNIST data on the deepexplain page:

line 94, in _set_check_baseline
if self.baseline is None:
AttributeError: 'GradientXInput' object has no attribute 'baseline'.

I know this is a long shot, but as I'm fairly frustrated, I would appreciate any input.

Also, since the bpnet is a ResNet structure, does this affect using deeplift/deepexplain to get contribution scores? Thanks so much!