Question

我正在omniglot数据集上构建暹罗模型。但是要避免使用Numpy，因为我希望模型尽可能通用。并且可以Git Repository。但是，我得到标题中描述的错误。当我尝试拟合模型时发生错误。下面的代码和下面的完整堆栈跟踪：

这是脚本代码

' Load Omniglot dataset ' 
ds, ds_info = tfds.load(name='Omniglot', with_info=True, as_supervised=True)

' Split the dataset into testing and training '
ds_train, ds_test = ds["train"], ds["test"]

'Create a CNN encoder'
model = conv_net (ds_info)

optimizer = keras.optimizers.Adam() 

' Compile the model with the contrastive loss function '
cont_loss_model = compile_cnn(model, contrastive_loss, optimizer)

' Fit the results '
cont_loss_model.fit(
    ds_train,
    validation_data=ds_test,
    epochs=epochs,
    verbose=1
    )

这里是使用的功能

@tf.function
def contrastive_loss(label, embedding, margin = 0.4):
    '''
    contrastive_loss function
    
    @param p: Positive vector
    @param n: Negetive vector
    
    @returns (float): y - Integer value representing distance
    '''
    # Assign the label
    y = label
    
    # Assign the embeddings
    p1 = embedding[0]
    p2 = embedding[1]
    
    # Get the euclean distance    
    d = tf.norm(p1 - p2, axis=-1)
    
    
    if y == 0:
        return (1/2) * tf.math.sqrt(d)
    else:
        return (1/2) * tf.math.sqrt(tf.math.maximum(0.0, (margin-d)))


#- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 

def conv_net (ds_info, batch_size = 128, epochs = 12):
'''
Create a cnn 

@param x    
@param y    
@param ds_info    
@optional loss    
@optional batch    
@optional opochs

'''

  # Get the input shape
image_shape = ds_info.features['image'].shape

model = keras.Sequential(
[
    Conv2D(32, 3, activation='relu', input_shape=image_shape),
    MaxPooling2D(),
    Conv2D(32, 3, activation='relu'),
    MaxPooling2D(),
    Flatten(),
    Dense(
        128, 
        activation='relu', 
        kernel_regularizer=regularizers.l2(0.01),
        bias_regularizer=regularizers.l1(0.01)
        ),
    Dense(ds_info.features['label'].num_classes, activation='softmax')
])

model.summary()

return model

这是错误

Input 0 of layer sequential_17 is incompatible with the layer: : expected min_ndim=4, found ndim=3. Full shape received: [105, 105, 3]

完整堆栈跟踪

C:\Users\User\anaconda3\lib\site-packages\tensorflow\python\keras\engine\training.py:806 train_function  *
    return step_function(self, iterator)
C:\Users\User\anaconda3\lib\site-packages\tensorflow\python\keras\engine\training.py:796 step_function  **
    outputs = model.distribute_strategy.run(run_step, args=(data,))
C:\Users\User\anaconda3\lib\site-packages\tensorflow\python\distribute\distribute_lib.py:1211 run
    return self._extended.call_for_each_replica(fn, args=args, kwargs=kwargs)
C:\Users\User\anaconda3\lib\site-packages\tensorflow\python\distribute\distribute_lib.py:2585 call_for_each_replica
    return self._call_for_each_replica(fn, args, kwargs)
C:\Users\User\anaconda3\lib\site-packages\tensorflow\python\distribute\distribute_lib.py:2945 _call_for_each_replica
    return fn(*args, **kwargs)
C:\Users\User\anaconda3\lib\site-packages\tensorflow\python\keras\engine\training.py:789 run_step  **
    outputs = model.train_step(data)
C:\Users\User\anaconda3\lib\site-packages\tensorflow\python\keras\engine\training.py:747 train_step
    y_pred = self(x, training=True)
C:\Users\User\anaconda3\lib\site-packages\tensorflow\python\keras\engine\base_layer.py:975 __call__
    input_spec.assert_input_compatibility(self.input_spec, inputs,
C:\Users\User\anaconda3\lib\site-packages\tensorflow\python\keras\engine\input_spec.py:191 assert_input_compatibility
    raise ValueError('Input ' + str(input_index) + ' of layer ' +

层顺序需要1个输入，但在仅使用张量模型时，它会收到4个输入张量

0 个答案: