attributeError：＆＃x27; dtype＆＃x27;对象没有属性＆＃x27; type＆＃x27;

Question

当我收到培训会产生损失和准确性（它们都是张量）时，我想使用tf.compat.v1.keras.utils.progbar（）功能和更新方法来更新它们，以便我可以得到更好查看训练过程，但是发生以下错误：

 File "E:/python/maml_mnist/train.py", line 64, in <module>
,('accuracy',mean_train_meta_acc)]
File "C:\Users\12130\AppData\Roaming\Python\Python36\site- 
packages\tensorflow\python\keras\utils\generic_utils.py", line 442, in update
avg = np.mean(self._values[k][0] / max(1, self._values[k][1]))
File "<__array_function__ internals>", line 6, in mean
File "E:\python\lib\site-packages\numpy\core\fromnumeric.py", line 3373, in mean
out=out, **kwargs)
File "E:\python\lib\site-packages\numpy\core\_methods.py", line 170, in _mean
ret = ret.dtype.type(ret / rcount)
AttributeError: 'DType' object has no attribute 'type' here

代码如下：

train_progbar = tf.compat.v1.keras.utils.Progbar(train_data.steps)
for i in range(train_data.steps):
    batch_train_loss,batch_train_acc = maml.train_on_batch(train_data.get_one_batch(),
                                                               inner_optimizer,
                                                               inner_step=1,
                                                               outer_optimizer=outer_optimizer)

    train_meta_loss.append(batch_train_loss)
    train_meta_acc.append(batch_train_acc)
    mean_train_meta_loss = np.mean(train_meta_loss)
    mean_train_meta_acc = np.mean(batch_train_acc)

    train_progbar.update(i+1
                             ,[('loss',mean_train_meta_loss)
                            ,('accuracy',mean_train_meta_acc)]
                             )

我知道错误是由于解决平均值而引起的，但我不知道如何修复它以成功运行代码。

我的TensorFlow版本是1.14.0。