Question

所以，我有一组大小为32x32x32的MR图像。我加入了一个转换网络。然而，我在最后一层中得到一个错误，如下所示： ValueError：检查模型目标时出错：预期activation_8有形状（None，1）但是有形状的数组（3,24）我花了很多时间试图调试这个错误，但是没有成功。

我的代码的关键如下所示。我使用train_on_batch使用批量大小为24来训练数据。任何关于出错的指示都将非常感激。

    def create_network():
      print 'Setting up first layer'
      model = Sequential()
      model.add(Convolution2D(32,3,3, border_mode = 'valid', subsample 
      =      (1,1), init = 'glorot_uniform',input_shape = 
     (32,32,32),dim_ordering = 'th'))
     model.add(BatchNormalization(epsilon=0.001, mode=2, axis=1, 
     momentum=0.99, weights=None, beta_init='zero', gamma_init='one', 
     gamma_regularizer=None, beta_regularizer=None))
     model.add(Activation('relu'))
     print 'First layer setup'

     print 'Setting up second layer'
     model.add(Convolution2D(32,3,3,border_mode = 'same', subsample = 
     (1,1),init = 'glorot_uniform'))
     model.add(BatchNormalization(epsilon=0.001, mode=2, axis=1, 
     momentum=0.99, weights=None, beta_init='zero', gamma_init='one', 
     gamma_regularizer=None, beta_regularizer=None))
     model.add(Activation('relu'))
     model.add(MaxPooling2D(pool_size = (2,2),strides = None))

     print 'Setting up third layer' 
     model.add(Convolution2D(64,3,3, border_mode = 'same', subsample = 
     (1,1), init = 'glorot_uniform'))
     model.add(BatchNormalization(epsilon=0.001, mode=2, axis=1, 
     momentum=0.99, weights=None, beta_init='zero', gamma_init='one', 
     gamma_regularizer=None, beta_regularizer=None))
     model.add(Activation('relu'))

     model.add(Convolution2D(64,3,3, border_mode = 'same', subsample =  
     1,1), init = 'glorot_uniform'))
     model.add(BatchNormalization(epsilon=0.001, mode=2, axis=1, 
     momentum=0.99, weights=None, beta_init='zero', gamma_init='one', 
     gamma_regularizer=None, beta_regularizer=None))
     model.add(Activation('relu'))
     model.add(MaxPooling2D(pool_size = (2,2),strides = None))

     model.add(Convolution2D(32,3,3,border_mode = 'same',subsample = 
     (1,1),init = 'glorot_uniform'))
     model.add(BatchNormalization(epsilon=0.001, mode=2, axis=1, 
     momentum=0.99, weights=None, beta_init='zero', gamma_init='one', 
     gamma_regularizer=None, beta_regularizer=None))
     model.add(Activation('relu'))

     model.add(Convolution2D(32,3,3,border_mode = 'same',subsample = 
     (1,1),init = 'glorot_uniform'))
     model.add(BatchNormalization(epsilon=0.001, mode=2, axis=1, 
     momentum=0.99, weights=None, beta_init='zero', gamma_init='one', 
     gamma_regularizer=None, beta_regularizer=None))
     model.add(Activation('relu'))
     model.add(MaxPooling2D(pool_size = (2,2),strides = None))

     model.add(Flatten()) 
     model.add(Dense(256))
     model.add(Activation('relu'))
     model.add(Dense(1))
     model.add(Activation('sigmoid'))

     sgd = SGD(lr=1e-8, decay=1e-6, momentum=0.9, nesterov=True)
     model.compile(optimizer = 'sgd',loss 
     ='binary_crossentropy',metrics = ['accuracy'])
     return model

    if __name__ == "__main__":
       model = create_network()
       batch_size = 24
       X_train,y_train = load_train_data()
       for i in range(300):
        model.train_on_batch(X_train,y_train)

最终的sigmoid层在Keras中给出了ValueError

0 个答案: