IDEA-CCNL
/

Erlangshen-TCBert-110M-Classification-Chinese

Model card Files Files and versions Community

idea-teacher commited on Oct 27, 2022

Commit

28264c7

•

1 Parent(s): 3a29994

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -31,7 +31,7 @@ The TCBert with 110M parameters is pre-trained for, not limited to, Chinese topi
 ## 模型信息 Model Information
-为了提高模型在句子匹配上的效果，我们收集了大量句子匹配数据进行预训练，随后在FewCLUE的BUSTM任务进行微调，所有的训练均基于我们提出的UniMC框架。最终结果表明，3.25亿参数的模型通过我们的训练策略可以在句子匹配任务上超过1.3亿参数的大模型。
 To improve the model performance on the topic classification task, we collected numerous topic classification datasets for pre-training based on general prompts.
 ### 下游效果 Performance

 ## 模型信息 Model Information
+为了提高模型在话题分类上的效果，我们收集了大量话题分类数据进行基于prompts的预训练。
 To improve the model performance on the topic classification task, we collected numerous topic classification datasets for pre-training based on general prompts.
 ### 下游效果 Performance