huggingface-transformers 快速上手bert 使用注意

你可能看过这篇：https://blog.csdn.net/qiguanjiezl/article/details/122504454

这个代码就能用，但是注意一点：不要自己下载离线model然后加载

 

因为：1. 你下载的可能版本不匹配

           2. 目前国内应该是有自动镜像，直接  model = BertForMaskedLM.from_pretrained('bert-base-uncased')   就行，下载速度很快。

 

可能提示：需要什么权限。

退出shell 或者 window cmd，选择sudo，或者window： 以管理员身份运行，打开 cmd窗口

-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

官网的例程：https://huggingface.co/docs/transformers/model_doc/bert#transformers.BertForMaskedLM

 

直接运行。注意结果需要print

print(tokenizer.decode(predicted_token_id))

 

可能打印：

只是info不是error，自己翻译一下就行了

Some weights of the model checkpoint at bert-base-uncased were not used when initializing BertForMaskedLM: ['cls.seq_relationship.weight', 'cls.seq_relationship.bias']
- This IS expected if you are initializing BertForMaskedLM from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).
- This IS NOT expected if you are initializing BertForMaskedLM from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).

 

最后给新手推荐一个讲解比较好的：https://www.bilibili.com/video/BV1Ao4y1k7AX/