标签结果预测问题

macanv / BERT-BiLSTM-CRF-NER

Tensorflow solution of NER task Using BiLSTM-CRF model with Google BERT Fine-tuning And private Server services

https://github.com/macanv/BERT-BiLSMT-CRF-NER

4.68k stars 1.25k forks source link

Closed Nalq11 closed 4 years ago

Nalq11 commented 4 years ago

博主，你好，想求教一下：现在使用自己的标签进行训练会遇到这样的问题：在训练小于三轮或者dropout等于0.5的时候预测比较正常，如果增加训练轮数或者调低dropout，预测会大幅度得增加B和I的概率，最后会全部变成B或者I。是自己的训练数据有问题么？