解决训练模型时内存泄漏的问题

class OxfordPets(keras.utils.Sequence):

    # 在__init__方法中指定batch_size,img_size,input_img_paths,target_img_paths
    def __init__(self, batch_size, img_size, input_img_paths, target_img_paths,shuffle=True):
        self.batch_size = batch_size  # 批量大小
        self.img_size = img_size  # 图像大小
        self.input_img_paths = input_img_paths  # 输入图像路径
        self.target_img_paths = target_img_paths  # 标注图像路径
        self.shuffle = shuffle
        self.on_epoch_end()

    def __len__(self):
        # 计算迭代次数
        return len(self.target_img_paths) // self.batch_size

    def __getitem__(self, idx):
        """
        获取每一个batch数据
        """
        i = idx * self.batch_size
        # 获取输入的图像数据
        batch_input_img_paths = self.input_img_paths[i: i + self.batch_size]
        # 获取标签数据
        batch_target_img_paths = self.target_img_paths[i: i + self.batch_size]
        # 构建特征值数据：获取图像数据中每个像素的数据存储在x中
        x = np.zeros((self.batch_size,) + self.img_size + (8,), dtype="float32")
        #x = np.zeros((batch_size,) + self.img_size + (1,), dtype="float32")
        for j, path in enumerate(batch_input_img_paths):
            #img = load_img(path, target_size=self.img_size)
            img = np.load(path)['arr_0']
            img = np.array(img)
            #x[j] = img/65535
            x[j] = img
        # 构建目标值数据：获取标注图像中每个像素中的数据存在y中
        y = np.zeros((self.batch_size,) + self.img_size +(1,), dtype="int32")
        for j, path in enumerate(batch_target_img_paths):
            #img =  cv2.imread(path,2)
            img = np.load(path)['arr_0']
            #y[j] = np.array(img)
            y[j] = np.expand_dims(img, 2)
        return x, y
    def on_epoch_end(self):
        self.indexes = np.arange(len(self.input_img_paths))
        if self.shuffle == True:
            np.random.shuffle(self.indexes)

input_dir = "./data/"
input_img_paths = sorted(
     [
          os.path.join(input_dir, fname)
          for fname in os.listdir(input_dir)
          if fname.endswith(".npz")
     ]
)