TypeError: reduce_sum() got an unexpected keyword argument 'keepdims' about capsnet-tensorflow

[vidzrox]

It should be written as "keep_dims=True"

[ZhiboRao]

I have met the same question.

[ZhiboRao]

I find you can fix the capsLayer.py. In 129, you delete the "keepdims=True".
Then it works.

[pulins201]

ok

[sks4world]

I'm getting the same error when following this github code
https://github.com/akosiorek/akosiorek.github.io/blob/master/notebooks/attention_glimpse.ipynb How can I fix this? I'm stuck

Jupyter notebook:

Gaussian Attention

gaussian_att_params = tf.concat([tu, ts, td, tu, ts, td], -1)
gaussian_glimpse_expr = gaussian_glimpse(tx, gaussian_att_params, glimpse_size)

Upon running error:
TypeError Traceback (most recent call last)
in
1 # Gaussian Attention
2 gaussian_att_params = tf.concat([tu, ts, td, tu, ts, td], -1)
----> 3 gaussian_glimpse_expr = gaussian_glimpse(tx, gaussian_att_params, glimpse_size)

in gaussian_glimpse(img_tensor, transform_params, crop_size)
29 uy, sy, dy, ux, sx, dx = tf.split(transform_params, 6, split_ax)
30 # create Gaussian masks, one for each axis
---> 31 Ay = gaussian_mask(uy, sy, dy, h, H)
32 Ax = gaussian_mask(ux, sx, dx, w, W)
33 # extract glimpse

in gaussian_mask(u, s, d, R, C)
14 mask = tf.exp(-.5 * tf.square(column_centres / s))
15 # we add eps for numerical stability
---> 16 normalised_mask = mask / (tf.reduce_sum(mask, 1, keep_dims=True) + 1e-8)
17 return normalised_mask
18

~\Anaconda2\envs\PythonCPU\lib\site-packages\tensorflow_core\python\util\dispatch.py in wrapper(*args, **kwargs)
178 """Call target, and fall back on dispatchers if there is a TypeError."""
179 try:
--> 180 return target(*args, **kwargs)
181 except (TypeError, ValueError):
182 # Note: convert_to_eager_tensor currently raises a ValueError, not a

TypeError: reduce_sum() got an unexpected keyword argument 'keep_dims'