Update README.md

README.md

@@ -23,7 +23,7 @@ widget:
 
 2021年登顶FewCLUE和ZeroCLUE的中文BERT，在数个改写任务微调后的版本
 
-This is the fine-tuned version of the Chinese BERT model on several semantic matching and similarity datasets, which topped FewCLUE and ZeroCLUE benchmark in 2021
+This is the fine-tuned version of the Chinese BERT model on several paraphrase datasets, which topped FewCLUE and ZeroCLUE benchmark in 2021
 
 ## 模型分类 Model Taxonomy
 
@@ -32,10 +32,10 @@ This is the fine-tuned version of the Chinese BERT model on several semantic mat
 | 通用 General  | 自然语言理解 NLU | 二郎神 Erlangshen | MegatronBert |      1.3B      |     Similarity     |
 
 ## 模型信息 Model Information
-We collect 20  paraphrace datasets in the Chinese domain for finetune, with a total of 2773880 samples
+
 基于[Erlangshen-MegatronBert-1.3B](https://huggingface.co/IDEA-CCNL/Erlangshen-MegatronBert-1.3B)，我们在收集的20个用于finetune的中文领域的改写数据集，总计227347个样本上微调了一个Similarity版本。
 
-Based on [Erlangshen-MegatronBert-1.3B] (https://huggingface.co/IDEA-CCNL/Erlangshen-MegatronBert-1.3B), we fine-tuned a similarity version on 8 Chinese paraphrace datasets, with totaling 227,347 samples.
+Based on [Erlangshen-MegatronBert-1.3B] (https://huggingface.co/IDEA-CCNL/Erlangshen-MegatronBert-1.3B), we fine-tuned a similarity version on 8 Chinese paraphrase datasets, with totaling 227,347 samples.
 
 ### 下游效果 Performance