处理多个npy图像阵列时使用Keras DataGenerator时出错

Question

我试图实现一个回归问题。我的输入文件是446个文件，每个文件包含6k到12k大小的帧（150150,3），例如input316是（6832150150,3），446个文件具有相应的长度和51个维度（整数值），例如output316是（6832,51）。由于整个数据集的容量太大，无法放入RAM中，因此我决定使用DataGenerator类keras.utils.Sequence。我按照这个link的指示，将dim参数编辑为（150150），将n_通道编辑为3。这是我的最终代码：

    'Generates data for Keras'
    def __init__(self, list_IDs, batch_size=32, dim=(150,150), 
                 n_channels=3, shuffle=True):
        'Initialization'
        self.dim = dim
        self.batch_size = batch_size
        self.list_IDs = list_IDs
        self.n_channels = n_channels
        self.shuffle = shuffle
        self.on_epoch_end()

    def __len__(self):
        'Denotes the number of batches per epoch'
        return int(np.floor(len(self.list_IDs) / self.batch_size))

    def __getitem__(self, index):
        'Generate one batch of data'
        # Generate indexes of the batch
        indexes = self.indexes[index*self.batch_size:(index+1)*self.batch_size]

        # Find list of IDs
        list_IDs_temp = [self.list_IDs[k] for k in indexes]

        # Generate data
        X, y = self.__data_generation(list_IDs_temp)

        return X, y

    def on_epoch_end(self):
        'Updates indexes after each epoch'
        print("Epoch is finished!")
        self.indexes = np.arange(len(self.list_IDs))
        if self.shuffle == True:
            np.random.shuffle(self.indexes)

    def __data_generation(self, list_IDs_temp):
        'Generates data containing batch_size samples' # X : (n_samples, *dim, n_channels)
        # Initialization
        X = np.empty((self.batch_size, *self.dim, self.n_channels))
        print('shape ='+ str(X.shape))
        y = np.empty((self.batch_size,51))

        # Generate data
        for i, ID in enumerate(list_IDs_temp):
            # Store sample
            X[i,] = np.load('Train/X' + str(ID) + '.npy')
            print('shapezz ='+ str(X.shape))
            # Store class
            y[i,] = np.load('Train/y' + str(ID) + '.npy')

但是，在尝试使用以下方法拟合模型时：

partition = [i for i in range(446)]
np.random.shuffle(partition)
training_generator = DataGenerator(partition)
model.fit_generator(generator=training_generator,
                    #use_multiprocessing=True,
                    #workers=4,
                    verbose = 1)

我面对这个错误：

ValueError: could not broadcast input array from shape (6145,150,150,3) into shape (150,150,3)

有人知道我怎么解决这个问题吗？代码的哪一部分以及如何更改？提前谢谢