example->ner->ner.py always returns text='...' and start=0

I use the ml-backend and the pretrained bert-base-chinese model to train a NER model，The model is trained ok, but when I use it to predict, it always returns text='...' and start=0, in fact, the string doesn't contain any '...' at all. When i look into the code, it shows the return text is set to '...'

What's wrong?

Any help is appreciated！

Start with example-ner-ner.py, where the issue reports that the prediction result sets text to '...' and start to 0. Trace how the trained model's prediction is converted into the returned entity and compare it with the input string. Done means the returned text and start offset identify the actual entity span.

Written by the indexing model from the issue text.