ValueError: Dimensions must be equal, but are 2560 and 10 for 'digit_caps_2/add' (op: 'AddV2') with input shapes: [?,10,10,2560,32], [10,32].

[QuickLearner171998]

I am facing this error while training. I have used the following args: class args: numGPU = 1 epochs = 100 batch_size = 8 lr = 0.001 lr_decay = 0.96 lam_recon = 0.4 r = 3 routings = 3 shift_fraction = 0.1 debug = False digit = 5 save_dir = 'model/CIFAR10/13' t = False w = None ep_num = 0 dataset = "CIFAR10"

The error is common to all datasets.

[QuickLearner171998]

This is the exact error:

InvalidArgumentError: Dimensions must be equal, but are 2560 and 10 for 'digit_caps_2/add' (op: 'AddV2') with input shapes: [?,10,10,2560,32], [10,32].

During handling of the above exception, another exception occurred:

ValueError Traceback (most recent call last) 12 frames

in () 118 119 # model, eval_model = DeepCapsNet(input_shape=x_train.shape[1:], n_class=y_train.shape[1], routings=args.routings) # for 64*64 --> 120 model, eval_model = DeepCapsNet28(input_shape=x_train.shape[1:], n_class=y_train.shape[1], routings=args.routings) #for 28*28 121 122 # plot_model(model, show_shapes=True,to_file=args.save_dir + '/model.png') in DeepCapsNet28(input_shape, n_class, routings) 132 133 # l = Dropout(0.4)(l) --> 134 digits_caps = CapsuleLayer(num_capsule=n_class, dim_capsule=32, routings=routings, channels=0, name='digit_caps')(l) 135 136 l = CapsToScalars(name='capsnet')(digits_caps) /usr/local/lib/python3.6/dist-packages/keras/engine/base_layer.py in __call__(self, inputs, **kwargs) 449 # Actually call the layer, 450 # collecting output(s), mask(s), and shape(s). --> 451 output = self.call(inputs, **kwargs) 452 output_mask = self.compute_mask(inputs, previous_mask) 453 in call(self, inputs, training) 559 # then matmal: [input_num_capsule] x [input_num_capsule, dim_capsule] -> [dim_capsule]. 560 # outputs.shape=[None, num_capsule, dim_capsule] --> 561 outputs = squash(K.batch_dot(c, inputs_hat, [2, 2]) + self.B) # [None, 10, 16] 562 563 if i < self.routings - 1: /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/math_ops.py in binary_op_wrapper(x, y) 910 else: 911 raise --> 912 return func(x, y, name=name) 913 914 def binary_op_wrapper_sparse(sp_x, y): /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/math_ops.py in _add_dispatch(x, y, name) 1195 return gen_math_ops.add(x, y, name=name) 1196 else: -> 1197 return gen_math_ops.add_v2(x, y, name=name) 1198 else: 1199 return gen_math_ops.add(x, y, name=name) /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/gen_math_ops.py in add_v2(x, y, name) 547 # Add nodes to the TensorFlow graph. 548 _, _, _op = _op_def_lib._apply_op_helper( --> 549 "AddV2", x=x, y=y, name=name) 550 _result = _op.outputs[:] 551 _inputs_flat = _op.inputs /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/op_def_library.py in _apply_op_helper(self, op_type_name, name, **keywords) 792 op = g.create_op(op_type_name, inputs, dtypes=None, name=scope, 793 input_types=input_types, attrs=attr_protos, --> 794 op_def=op_def) 795 796 # Conditionally invoke tfdbg v2's op callback(s). /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/util/deprecation.py in new_func(*args, **kwargs) 505 'in a future version' if date is None else ('after %s' % date), 506 instructions) --> 507 return func(*args, **kwargs) 508 509 doc = _add_deprecated_arg_notice_to_docstring( /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/ops.py in create_op(***failed resolving arguments***) 3355 raise TypeError("Input #%d is not a tensor: %s" % (idx, a)) 3356 return self._create_op_internal(op_type, inputs, dtypes, input_types, name, -> 3357 attrs, op_def, compute_device) 3358 3359 def _create_op_internal( /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/ops.py in _create_op_internal(self, op_type, inputs, dtypes, input_types, name, attrs, op_def, compute_device) 3424 input_types=input_types, 3425 original_op=self._default_original_op, -> 3426 op_def=op_def) 3427 self._create_op_helper(ret, compute_device=compute_device) 3428 return ret /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/ops.py in __init__(self, node_def, g, inputs, output_types, control_inputs, input_types, original_op, op_def) 1768 op_def, inputs, node_def.attr) 1769 self._c_op = _create_c_op(self._graph, node_def, grouped_inputs, -> 1770 control_input_ops) 1771 # pylint: enable=protected-access 1772 /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/ops.py in _create_c_op(graph, node_def, inputs, control_inputs) 1608 except errors.InvalidArgumentError as e: 1609 # Convert to ValueError for backwards compatibility. -> 1610 raise ValueError(str(e)) 1611 1612 return c_op ValueError: Dimensions must be equal, but are 2560 and 10 for 'digit_caps_2/add' (op: 'AddV2') with input shapes: [?,10,10,2560,32], [10,32].

[brjathu]

For CIFAR10, you need to uncomment line 122 in train.py and comment line 123 in train.py.

[christopher-lw]

I used the exact train.py file from this repository and ended up in a similar error. I'm using Google Colab with GPU hardware accelerator. The tensorflow version used is 1.15.0.

This is the error message: Traceback (most recent call last): File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/ops.py", line 1607, in _create_c_op c_op = c_api.TF_FinishOperation(op_desc) tensorflow.python.framework.errors_impl.InvalidArgumentError: Dimensions must be equal, but are 640 and 10 for 'digit_caps/add' (op: 'AddV2') with input shapes: [?,10,10,640,32], [10,32].

During handling of the above exception, another exception occurred:

Traceback (most recent call last): File "train.py", line 123, in model, eval_model = DeepCapsNet28(input_shape=x_train.shape[1:], n_class=y_train.shape[1], routings=args.routings) #for 28*28 File "/content/deepcaps/deepcaps.py", line 135, in DeepCapsNet28 digits_caps = CapsuleLayer(num_capsule=n_class, dim_capsule=32, routings=routings, channels=0, name='digit_caps')(l) File "/usr/local/lib/python3.6/dist-packages/keras/engine/base_layer.py", line 451, in call output = self.call(inputs, *kwargs) File "/content/deepcaps/capslayers.py", line 563, in call outputs = squash(K.batch_dot(c, inputs_hat, [2, 2]) + self.B) # [None, 10, 16] File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/math_ops.py", line 912, in binary_op_wrapper return func(x, y, name=name) File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/math_ops.py", line 1197, in _add_dispatch return gen_math_ops.add_v2(x, y, name=name) File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/gen_math_ops.py", line 549, in add_v2 "AddV2", x=x, y=y, name=name) File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/op_def_library.py", line 794, in _apply_op_helper op_def=op_def) File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/util/deprecation.py", line 507, in new_func return func(args, **kwargs) File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/ops.py", line 3357, in create_op attrs, op_def, compute_device) File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/ops.py", line 3426, in _create_op_internal op_def=op_def) File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/ops.py", line 1770, in init control_input_ops) File "/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/framework/ops.py", line 1610, in _create_c_op raise ValueError(str(e)) ValueError: Dimensions must be equal, but are 640 and 10 for 'digit_caps/add' (op: 'AddV2') with input shapes: [?,10,10,640,32], [10,32].

[brjathu]

you need to uncomment line 122 in train.py and comment line 123 in train.py. This is causing the error, DeepCapsNet28 is for MNIST, use DeepCapsNet for CIFAR10.

[mrazekv]

Hi @brjathu I got exactly the same error. I am using DeepCapsNet28 for MNIST dataset (however I tried to change toggle comments on lines 122 and 123 without any impact).

I have tested several environments and no one was working. Can you please help me with this error? I think it may be caused by Keras update.

Thanks in advance!

This is the trackback of the error

WARNING:tensorflow:From $HOME/deepcaps/deepcaps/capslayers.py:346: calling softmax (from tensorflow.python.ops.nn_ops) with dim is deprecated and will be removed in a future version.
Instructions for updating:
dim is deprecated, use axis instead
Traceback (most recent call last):
  File "$HOME/.conda/envs/snowflakes/lib/python3.6/site-packages/tensorflow/python/framework/ops.py", line 1659, in _create_c_op
    c_op = c_api.TF_FinishOperation(op_desc)
tensorflow.python.framework.errors_impl.InvalidArgumentError: Dimensions must be equal, but are 640 and 10 for 'digit_caps/add' (op: 'Add') with input shapes: [?,10,10,640,32], [10,32].

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "train.py", line 123, in <module>
    model, eval_model = DeepCapsNet28(input_shape=x_train.shape[1:], n_class=y_train.shape[1], routings=args.routings)  #for 28*28
  File "$HOME/deepcaps/deepcaps/deepcaps.py", line 135, in DeepCapsNet28
    digits_caps = CapsuleLayer(num_capsule=n_class, dim_capsule=32, routings=routings, channels=0, name='digit_caps')(l)
  File "$HOME/.local/lib/python3.6/site-packages/keras/engine/base_layer.py", line 489, in __call__
    output = self.call(inputs, **kwargs)
  File "$HOME/deepcaps/deepcaps/capslayers.py", line 563, in call
    outputs = squash(K.batch_dot(c, inputs_hat, [2, 2]) + self.B)  # [None, 10, 16]
  File "$HOME/.conda/envs/snowflakes/lib/python3.6/site-packages/tensorflow/python/ops/math_ops.py", line 824, in binary_op_wrapper
    return func(x, y, name=name)
  File "$HOME/.conda/envs/snowflakes/lib/python3.6/site-packages/tensorflow/python/ops/gen_math_ops.py", line 365, in add
    "Add", x=x, y=y, name=name)
  File "$HOME/.conda/envs/snowflakes/lib/python3.6/site-packages/tensorflow/python/framework/op_def_library.py", line 788, in _apply_op_helper
    op_def=op_def)
  File "$HOME/.conda/envs/snowflakes/lib/python3.6/site-packages/tensorflow/python/util/deprecation.py", line 507, in new_func
    return func(*args, **kwargs)
  File "$HOME/.conda/envs/snowflakes/lib/python3.6/site-packages/tensorflow/python/framework/ops.py", line 3300, in create_op
    op_def=op_def)
  File "$HOME/.conda/envs/snowflakes/lib/python3.6/site-packages/tensorflow/python/framework/ops.py", line 1823, in __init__
    control_input_ops)
  File "$HOME/.conda/envs/snowflakes/lib/python3.6/site-packages/tensorflow/python/framework/ops.py", line 1662, in _create_c_op
    raise ValueError(str(e))
ValueError: Dimensions must be equal, but are 640 and 10 for 'digit_caps/add' (op: 'Add') with input shapes: [?,10,10,640,32], [10,32].

This is the list of installed packages (I am using Anaconda)

$ conda list
# packages in environment at $HOME/.conda/envs/snowflakes:
#
# Name                    Version                   Build  Channel
_tflow_select             2.1.0                       gpu    anaconda
absl-py                   0.9.0                    py36_0    anaconda
astor                     0.8.0                    py36_0    anaconda
blas                      1.0                         mkl    anaconda
c-ares                    1.15.0            h7b6447c_1001    anaconda
ca-certificates           2020.1.1                      0    anaconda
certifi                   2019.11.28               py36_0    anaconda
cudatoolkit               10.0.130                      0    anaconda
cudnn                     7.6.5                cuda10.0_0    anaconda
cupti                     10.0.130                      0    anaconda
freetype                  2.9.1                h8a8886c_1    anaconda
gast                      0.3.3                      py_0    anaconda
grpcio                    1.27.2           py36hf8bcb03_0    anaconda
h5py                      2.10.0           py36h7918eee_0    anaconda
hdf5                      1.10.4               hb1b8bf9_0    anaconda
intel-openmp              2020.0                      166    anaconda
jpeg                      9b                   habf39ab_1    anaconda
keras-applications        1.0.8                      py_0    anaconda
keras-base                2.3.1                    py36_0    anaconda
keras-gpu                 2.1.6                    py36_0    anaconda
keras-preprocessing       1.1.0                      py_1    anaconda
ld_impl_linux-64          2.33.1               h53a641e_7    anaconda
libedit                   3.1.20181209         hc058e9b_0    anaconda
libffi                    3.2.1                h4deb6c0_3    anaconda
libgcc-ng                 9.1.0                hdf63c60_0    anaconda
libgfortran-ng            7.3.0                hdf63c60_0    anaconda
libpng                    1.6.37               hbc83047_0    anaconda
libprotobuf               3.11.4               hd408876_0    anaconda
libstdcxx-ng              9.1.0                hdf63c60_0    anaconda
libtiff                   4.1.0                h2733197_0    anaconda
markdown                  3.1.1                    py36_0    anaconda
mkl                       2019.5                      281    anaconda
mkl-service               2.3.0            py36he904b0f_0    anaconda
mkl_fft                   1.0.15           py36ha843d7b_0    anaconda
mkl_random                1.1.0            py36hd6b4f25_0    anaconda
mock                      4.0.1                      py_0    anaconda
ncurses                   6.2                  he6710b0_0    anaconda
numpy                     1.18.1           py36h4f9e942_0    anaconda
numpy-base                1.18.1           py36hde5b4d6_1    anaconda
olefile                   0.46                       py_0    anaconda
openssl                   1.1.1                h7b6447c_0    anaconda
pillow                    7.0.0            py36hb39fc2d_0    anaconda
pip                       20.0.2                   py36_1    anaconda
protobuf                  3.11.4           py36he6710b0_0    anaconda
python                    3.6.10               h0371630_0    anaconda
pyyaml                    5.3              py36h7b6447c_0    anaconda
readline                  7.0                  h7b6447c_5    anaconda
scipy                     1.4.1            py36h0b6359f_0    anaconda
setuptools                45.2.0                   py36_0    anaconda
six                       1.14.0                   py36_0    anaconda
sqlite                    3.31.1               h7b6447c_0    anaconda
tensorboard               1.13.1           py36hf484d3e_0    anaconda
tensorflow                1.13.1          gpu_py36h3991807_0    anaconda
tensorflow-base           1.13.1          gpu_py36h8d69cac_0    anaconda
tensorflow-estimator      1.13.0                     py_0    anaconda
tensorflow-gpu            1.13.1               h0d30ee6_0    anaconda
termcolor                 1.1.0                    py36_1    anaconda
tk                        8.6.8                hbc83047_0    anaconda
werkzeug                  1.0.0                      py_0    anaconda
wheel                     0.34.2                   py36_0    anaconda
xz                        5.2.4                h14c3975_4    anaconda
yaml                      0.1.7                h96e3832_1    anaconda
zlib                      1.2.11               h7b6447c_3    anaconda
zstd                      1.3.7                h0b5b093_0    anaconda

[mrazekv]

Maybe I have found the problem!

Problem was in keras versions. With following modules

keras-base                2.2.4                    py36_0
keras-gpu                 2.2.4                         0
keras-preprocessing       1.1.0                      py_1    anaconda

and update in capslayers.py

#        self.data_format = K.normalize_data_format(data_format)
        self.data_format = conv_utils.normalize_data_format(data_format)

it seems to be working!

I recommend to update README with specification of keras version (in conda install commands) and this update in capslayers.py. I can create a pull request, if you want ;)

[brjathu]

Hi @mrazekv, First of all, sorry for the late reply, and Thank you for figuring out the issue. It would be great if can create the pull request.