当我尝试为WGAN_GP建立对抗图时，出现此错误。
def _create_adversarial_graph(self):
real_img = Input(shape=self.input_dims)
latent_space = Input(shape=(self.latent_space_size,))
fake_img = self.generator(latent_space)
# interpolate real and fake images
alpha = K.random_uniform((self.batch_size, 1, 1, 1))
self.interpolated_img = Add()([Multiply()([alpha, real_img]), Multiply()([1-alpha, fake_img])])
# pass it through discriminator
real_critic = self.discriminator(real_img)
fake_critic = self.discriminator(fake_img)
interpolated_critic = self.discriminator(self.interpolated_img)
#--------------------------------------------
# discriminator (critic) computational graph
#--------------------------------------------
set_trainable(self.generator, False) # freeze weights for generator while training discriminator
self.discriminator_model = Model(inputs=[real_img, latent_space], outputs=[real_critic, fake_critic, interpolated_critic])
self.discriminator_model.compile(
loss=[self.Wasserstein_loss, self.Wasserstein_loss, self.GP_loss],
optimizer=self.optimizer,
loss_weights=self.discriminator_loss_weights
)
self.generator和self.discriminator也是Keras模型。
完整错误：
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
<ipython-input-35-66b73f0e6b92> in <module>()
24 generator_padding=["same","same","same","same"],
25 optimizer=OPTIMIZER,
---> 26 batch_size=BATCH_SIZE
27 )
28 print("discriminator_model and generator_model summary:")
10 frames
<ipython-input-34-1ccb34bddd33> in __init__(self, input_dims, latent_space_size, discriminator_filters, discriminator_kernel_size, discriminator_strides, discriminator_padding, discriminator_loss_weights, generator_init_dense, generator_batch_norm_momentum, generator_filters, generator_kernel_size, generator_strides, generator_padding, optimizer, batch_size)
40 self.batch_size = batch_size
41 self.discriminator_loss_weights = discriminator_loss_weights
---> 42 self._create_model()
43 def _create_model(self):
44 self._create_discriminator()
<ipython-input-34-1ccb34bddd33> in _create_model(self)
45 self._create_generator()
46 # now that we have generator and discriminator we can make computational graph
---> 47 self._create_adversarial_graph()
48
49 def _create_discriminator(self):
<ipython-input-34-1ccb34bddd33> in _create_adversarial_graph(self)
98 #--------------------------------------------
99 set_trainable(self.generator, False) # freeze weights for generator while training discriminator
--> 100 self.discriminator_model = Model(inputs=[real_img, latent_space], outputs=[real_critic, fake_critic, interpolated_critic])
101 self.discriminator_model.compile(
102 loss=[self.Wasserstein_loss, self.Wasserstein_loss, self.GP_loss],
/tensorflow-1.15.2/python3.6/keras/legacy/interfaces.py in wrapper(*args, **kwargs)
89 warnings.warn('Update your `' + object_name + '` call to the ' +
90 'Keras 2 API: ' + signature, stacklevel=2)
---> 91 return func(*args, **kwargs)
92 wrapper._original_function = func
93 return wrapper
/tensorflow-1.15.2/python3.6/keras/engine/network.py in __init__(self, *args, **kwargs)
92 'inputs' in kwargs and 'outputs' in kwargs):
93 # Graph network
---> 94 self._init_graph_network(*args, **kwargs)
95 else:
96 # Subclassed network
/tensorflow-1.15.2/python3.6/keras/engine/network.py in _init_graph_network(self, inputs, outputs, name, **kwargs)
239 # Keep track of the network's nodes and layers.
240 nodes, nodes_by_depth, layers, layers_by_depth = _map_graph_network(
--> 241 self.inputs, self.outputs)
242 self._network_nodes = nodes
243 self._nodes_by_depth = nodes_by_depth
/tensorflow-1.15.2/python3.6/keras/engine/network.py in _map_graph_network(inputs, outputs)
1432 layer=layer,
1433 node_index=node_index,
-> 1434 tensor_index=tensor_index)
1435
1436 for node in reversed(nodes_in_decreasing_depth):
/tensorflow-1.15.2/python3.6/keras/engine/network.py in build_map(tensor, finished_nodes, nodes_in_progress, layer, node_index, tensor_index)
1419 tensor_index = node.tensor_indices[i]
1420 build_map(x, finished_nodes, nodes_in_progress, layer,
-> 1421 node_index, tensor_index)
1422
1423 finished_nodes.add(node)
/tensorflow-1.15.2/python3.6/keras/engine/network.py in build_map(tensor, finished_nodes, nodes_in_progress, layer, node_index, tensor_index)
1419 tensor_index = node.tensor_indices[i]
1420 build_map(x, finished_nodes, nodes_in_progress, layer,
-> 1421 node_index, tensor_index)
1422
1423 finished_nodes.add(node)
/tensorflow-1.15.2/python3.6/keras/engine/network.py in build_map(tensor, finished_nodes, nodes_in_progress, layer, node_index, tensor_index)
1419 tensor_index = node.tensor_indices[i]
1420 build_map(x, finished_nodes, nodes_in_progress, layer,
-> 1421 node_index, tensor_index)
1422
1423 finished_nodes.add(node)
/tensorflow-1.15.2/python3.6/keras/engine/network.py in build_map(tensor, finished_nodes, nodes_in_progress, layer, node_index, tensor_index)
1391 ValueError: if a cycle is detected.
1392 """
-> 1393 node = layer._inbound_nodes[node_index]
1394
1395 # Prevent cycles.
AttributeError: 'NoneType' object has no attribute '_inbound_nodes'