I'm using TF 2.0, and I get this error when I import tangent, due to a list of non-differentiable functions that includes tf.to_float (line 60), which is deprecated:

https://www.tensorflow.org/versions/r1.14/api_docs/python/tf/to_float

Okay, so this might not exactly be a "good first issue" - it is a little more advanced, but is still very much accessible to newcomers.

Similar to the mygrad.nnet.max_pool function, I would like there to be a mean-pooling layer. That is, a convolution-style windows is strided over the input, an

e.g.

julia> check_equal((;a=2, b=2), (;a=2,c=2))
Test Passed

This is because it hits the general iterator fallback.

This isn't super high priority, since it only comes up in chaecking add!! behavour
(as neither value is a AbstractTangent subtype).

Still would be nice to fix.
Need to channel it towards one of our overloads for structured types