transformer模型训练时，没有训练就出现了StopIteration的错误
Created by: dashulu
按照wiki https://github.com/PaddlePaddle/models/blob/develop/fluid/PaddleNLP/neural_machine_translation/transformer/README_cn.md 执行transformer， 得到的日志为： 65 [2019-01-07 20:15:17,001 INFO train.py:608] epoch: 25, consumed 0 .000111s 67 [2019-01-07 20:15:19,531 INFO train.py:608] epoch: 26, consumed 0 .000145s 69 [2019-01-07 20:15:21,376 INFO train.py:608] epoch: 27, consumed 0 .000136s 71 [2019-01-07 20:15:22,998 INFO train.py:608] epoch: 28, consumed 0 .000079s 73 [2019-01-07 20:15:24,641 INFO train.py:608] epoch: 29, consumed 0 .000208s 并没有进行真正的训练，直接出现StopIteration的excetion。 请问这个是什么原因？