BERT-chinese-text-classification-pytorch

BERT-chinese-text-classification-pytorch copied to clipboard

1. readme中说 那些pretrained 文件放在 base-uncased. 需要在base.py 中改掉 config 里的路径base-chinese 2. 给的cnews，如果默认 gbk则不能读入，需要 在open(raw_data_path,'r')中加入,encoding='utf-8' 。这个应该是本地默认编码问题，不大 3. 运行train步骤，load预训练权重会提示 在 bert_for_multi_class.py", line 11 init_weights() takes 1 positional argument but 2 were given 找了一会，应该是...

jingyonglin

Which file encoding does the vocab.txt use? I open it with unreadable code and there is no chinese word.

sakerbug

You put a link for the cnews.txt, but where is the test.txt?

rqiao

deserialized_objects[key]._set_from_file(f, offset, f_should_read_directly) RuntimeError: storage has wrong size: expected 0 got 2359296

3


JohnParken

您好，感谢分享。请问方便分享下处理后的数据集吗？

tectal