Exercise 2 - convolutional_model 'Dense' object has no attribute op

Course Q&A Deep Learning Specialization Convolutional Neural Networks
1

in
----> 1 conv_model = convolutional_model((64, 64, 3))
2 conv_model.compile(optimizer=‘adam’,
3 loss=‘categorical_crossentropy’,
4 metrics=[‘accuracy’])
5 conv_model.summary()

in convolutional_model(input_shape)
39
40 # YOUR CODE ENDS HERE
—> 41 model = tf.keras.Model(inputs=input_img, outputs=outputs)
42 return model

/opt/conda/lib/python3.7/site-packages/tensorflow/python/keras/engine/training.py in new(cls, *args, **kwargs)
240 # Functional model
241 from tensorflow.python.keras.engine import functional # pylint: disable=g-import-not-at-top
→ 242 return functional.Functional(*args, **kwargs)
243 else:
244 return super(Model, cls).new(cls, *args, **kwargs)

/opt/conda/lib/python3.7/site-packages/tensorflow/python/training/tracking/base.py in _method_wrapper(self, *args, **kwargs)
455 self._self_setattr_tracking = False # pylint: disable=protected-access
456 try:
→ 457 result = method(self, *args, **kwargs)
458 finally:
459 self._self_setattr_tracking = previous_value # pylint: disable=protected-access

/opt/conda/lib/python3.7/site-packages/tensorflow/python/keras/engine/functional.py in init(self, inputs, outputs, name, trainable)
113 # ‘arguments during initialization. Got an unexpected argument:’)
114 super(Functional, self).init(name=name, trainable=trainable)
→ 115 self._init_graph_network(inputs, outputs)
116
117 @trackable.no_automatic_dependency_tracking

/opt/conda/lib/python3.7/site-packages/tensorflow/python/training/tracking/base.py in _method_wrapper(self, *args, **kwargs)
455 self._self_setattr_tracking = False # pylint: disable=protected-access
456 try:
→ 457 result = method(self, *args, **kwargs)
458 finally:
459 self._self_setattr_tracking = previous_value # pylint: disable=protected-access

/opt/conda/lib/python3.7/site-packages/tensorflow/python/keras/engine/functional.py in _init_graph_network(self, inputs, outputs)
140
141 if any(not hasattr(tensor, ‘_keras_history’) for tensor in self.outputs):
→ 142 base_layer_utils.create_keras_history(self._nested_outputs)
143
144 self._validate_graph_inputs_and_outputs()

/opt/conda/lib/python3.7/site-packages/tensorflow/python/keras/engine/base_layer_utils.py in create_keras_history(tensors)
189 the raw Tensorflow operations.
190 “”"
→ 191 _, created_layers = _create_keras_history_helper(tensors, set(), )
192 return created_layers
193

/opt/conda/lib/python3.7/site-packages/tensorflow/python/keras/engine/base_layer_utils.py in _create_keras_history_helper(tensors, processed_ops, created_layers)
224 'op wrapping. Please wrap these ops in a Lambda layer: ’
225 ‘\n\n\n{example}\n\n’.format(example=example))
→ 226 op = tensor.op # The Op that created this Tensor.
227 if op not in processed_ops:
228 # Recursively set _keras_history.

AttributeError: ‘Dense’ object has no attribute ‘op’

sorry may i ask what the error mean? thanks!

2

Please click my name and message the code for the function def convolutional_model

3

Please see this link.

When following the functional approach to create a model, 2 steps are involved for each layer.

  1. Layer creation.
  2. Calling the layer.

This code

x = tf.keras.layers.Dense(units=1000)(previous_layer)

is the same as:

x = tf.keras.layers.Dense(units=1000) # creates new object
x = x(previous_layer) # invokes __call__ method of that object.
4

HI balaji, may i ask what is the previous layer?

i pass input image to Z1 and pass Z1 to Z2 ,and treate Z2 as the input layer is correct?

5

tf.keras.Input is an exception where you don’t need to call the layer.

previous_layer refers to the layer whose output is fed to this layer.

Getting back to the function, here are some hints to fix the code:

input_img = Input(shape=...)
Z1 = Conv2D(...)(input_img)
A1 = ReLU() (Z1)
# keep going.

Does this help?

1 Like
6

Hi balaji,
i did what you said but still error here

[code removed - moderator]

may i ask any mistake i made? thanks in advance

7

You did the Conv2D layers correctly and the pooling logic, but you’ve made the same mistake as earlier on the ReLU layers and the Flatten layer. Those also require the 2 step process that Balaji described above. All Keras “Layer” functions require the two levels of invocation.