Skip to content

Commit

Permalink
修改readme
Browse files Browse the repository at this point in the history
  • Loading branch information
920232796 committed Nov 22, 2021
1 parent 9ea2e1a commit 69c8584
Showing 1 changed file with 1 addition and 1 deletion.
2 changes: 1 addition & 1 deletion README.md
Original file line number Diff line number Diff line change
Expand Up @@ -20,7 +20,7 @@
4. relation_extract 关系抽取,比如三元组抽取任务。(复现苏剑林老师的例子,不完全一样。)
5. simbert SimBert模型,生成相似句子,并且给相似句子进行相似度判断。

五种任务分别加载四种不同的模型,通过``` model_class="seq2seq" or "cls" or "sequence_labeling" or "sequence_labeling_crf" or "relation_extrac" or "simbert"``` 参数去设置。具体可以去看examples里面的各种例子。
加载不同模型通过设置“model_name”参数实现,不同任务通过设置“model_class”参数实现,具体可以去看examples里面的各种例子。
### 预训练模型下载地址汇总:
1. roberta模型,模型和字典文件需要去 https://drive.google.com/file/d/1iNeYFhCBJWeUsIlnW_2K6SMwXkM4gLb_/view 这里下载。 具体可以参考这个github仓库~ https://github.com/ymcui/Chinese-BERT-wwm ,roberta-large模型也是在里面进行下载即可。
2. bert模型(目前不支持large),下载bert中文预训练权重 "bert-base-chinese": "https://s3.amazonaws.com/models.huggingface.co/bert/bert-base-chinese-pytorch_model.bin", 下载bert中文字典 "bert-base-chinese": "https://s3.amazonaws.com/models.huggingface.co/bert/bert-base-chinese-vocab.txt".
Expand Down

0 comments on commit 69c8584

Please sign in to comment.