异常报错

Tensorflow常见的问题：

• Cannot interpret feed_dict key as Tensor:
    是因为feed_dict中的key在计算图中找不到。

sess.run(loss, feed_dict={"input_ids":inputs, "labels":labels})

  上面代码的错误在于将key"input_ids"写成了字符串，正确如下：

sess.run(loss, feed_dict={input_ids:inputs, labels:labels}

• Tensor Tensor(“Placeholder:0”, shape=(?, 250), dtype=int32) is not an element of this graph.
    这个是因为***把计算图写tf.Graph外面了，无语。

• Tensor objects are only iterable when eager execution is enabled. To iterate over this tensor use tf.map_fn.
    原因是tensor placeholder对象不可能迭代遍历.

    for single in input_ids:
        print("aaa\n")

• InternalError: Dst tensor is not initialized
    OOM的问题, 重启pycharm不好使, 直接重启电脑。