diff --git a/ernie-gram/README.en.md b/ernie-gram/README.en.md
index bb864b00b089e15021e3496b21e33eba2926d1b9..cec672dddbb5a178d06b89e19b5af423eec882e7 100644
--- a/ernie-gram/README.en.md
+++ b/ernie-gram/README.en.md
@@ -1,6 +1,6 @@
 English|[简体中文](./README.zh.md)
 
-`Remind`: *ERNIE-Gram* model has been officially released in [here](#3-download-pretrained-models-optional). Our reproduction codes will be released to [repro branch](https://github.com/PaddlePaddle/ERNIE/tree/repro) soon.
+`Remind`: *ERNIE-Gram* model has been officially released in [here](#3-download-pretrained-models-optional). Our reproduction codes have been released to [repro branch](https://github.com/PaddlePaddle/ERNIE/tree/repro/ernie-gram).
 
 
 ## _ERNIE-Gram_: Pre-Training with Explicitly N-Gram Masked Language Modeling for Natural Language Understanding
@@ -96,7 +96,7 @@ try eager execution with `dygraph model` :
 
  - See **ERNIE-Gram** paper [Appendix B.1-4](https://arxiv.org/abs/2010.12148)
 
-For full reproduction of paper results, please checkout to `repro` branch of this repo.
+For full reproduction of paper results, please checkout to `repro` branch of this repo, the site is at [ernie-gram](https://github.com/PaddlePaddle/ERNIE/tree/repro/ernie-gram).
 
 # Citation
 
diff --git a/ernie-gram/README.zh.md b/ernie-gram/README.zh.md
index 9744acac0941962539ea79c72277e46458575bda..3f2ff212a7defbf011b94fd496ba62cda42393cd 100644
--- a/ernie-gram/README.zh.md
+++ b/ernie-gram/README.zh.md
@@ -1,6 +1,6 @@
 [English](./README.en.md)|简体中文
 
-`提醒`: *ERNIE-Gram* 中/英文模型已经[正式开源](#3-下载预训练模型可选)，paper 复现代码也即将开源至 [repro分支](https://github.com/PaddlePaddle/ERNIE/tree/repro)。现在您可以使用基于 Paddle 2.0 全新升级、基于动静结合的新版 ERNIE 套件体验 *ERNIE-Gram* 中/英文开源模型。
+`提醒`: *ERNIE-Gram* 中/英文模型已经[正式开源](#3-下载预训练模型可选)，paper 复现代码开源至 [repro分支](https://github.com/PaddlePaddle/ERNIE/tree/repro/ernie-gram)。现在您可以使用基于 Paddle 2.0 全新升级、基于动静结合的新版 ERNIE 套件体验 *ERNIE-Gram* 中/英文开源模型。
 
 
 ## _ERNIE-Gram_: Pre-Training with Explicitly N-Gram Masked Language Modeling for Natural Language Understanding
@@ -115,7 +115,7 @@ data/xnli
 | MSRA-NER(SIGHAN2006) | 16      | 5e-5   | 10| 0.1 |
 
 
-若希望复现 paper 中的所有实验，请切换至本 repo 的 `repro` 分支。
+若希望复现 paper 中的所有实验，请切换至本 repo 的 `repro` 分支, 地址为[ernie-gram](https://github.com/PaddlePaddle/ERNIE/tree/repro/ernie-gram)。
 
 ### 文献引用