update style and desc

mao-wang · Mar 24, 2020 · b14422a · b14422a
1 parent 101789b
commit b14422a
Show file tree

Hide file tree

Showing 2 changed files with 15 additions and 22 deletions.
diff --git a/README.md b/README.md
@@ -23,7 +23,7 @@ Yiming Cui, Wanxiang Che, Ting Liu, Bing Qin, Ziqing Yang, Shijin Wang, Guoping
 ## 新闻
 2020/3/23 本目录发布的模型已接入[飞桨PaddleHub](https://github.com/PaddlePaddle/PaddleHub)，查看[快速加载](#快速加载)
 
-**2020/3/11 为了更好地了解需求，邀请您填写[调查问卷](https://wj.qq.com/s2/5637766/6281)，以便为大家提供更好的资源。**
+2020/3/11 为了更好地了解需求，邀请您填写[调查问卷](https://wj.qq.com/s2/5637766/6281)，以便为大家提供更好的资源。
 
 2020/2/26 哈工大讯飞联合实验室发布[知识蒸馏工具TextBrewer](https://github.com/airaria/TextBrewer)
 
@@ -45,9 +45,10 @@ Yiming Cui, Wanxiang Che, Ting Liu, Bing Qin, Ziqing Yang, Shijin Wang, Guoping
 |-|-|
 | [简介](#简介) | 介绍BERT-wwm基本原理 |
 | [中文模型下载](#中文模型下载) | 提供了BERT-wwm的下载地址 |
+| [快速加载](#快速加载) | 介绍了如何使用[🤗Transformers](https://github.com/huggingface/transformers)、[PaddleHub](https://github.com/PaddlePaddle/PaddleHub)快速加载模型 |
 | [模型对比](#模型对比) | 提供了本目录中模型的参数对比 |
 | [中文基线系统效果](#中文基线系统效果) | 列举了部分中文基线系统效果 |
-| [new 小参数量模型](#小参数量模型) | 列举了小参数量模型（3层Transformer）的效果 |
+| [小参数量模型](#小参数量模型) | 列举了小参数量模型（3层Transformer）的效果 |
 | [使用建议](#使用建议) | 提供了若干使用中文预训练模型的建议 |
 | [英文模型下载](#英文模型下载) | 谷歌官方的英文BERT-wwm下载地址 |
 | [FAQ](#FAQ) | 常见问题答疑 |
@@ -113,8 +114,8 @@ chinese_wwm_L-12_H-768_A-12.zip
 PyTorch版本则包含`pytorch_model.bin`, `bert_config.json`, `vocab.txt`文件。
 
 
-### 快速加载
-1. 使用Huggingface-Transformers
+## 快速加载
+### 使用Huggingface-Transformers
 
 依托于[Huggingface-Transformers 2.2.2](https://github.com/huggingface/transformers)，可轻松调用以上模型。
 ```
@@ -134,9 +135,9 @@ model = BertModel.from_pretrained("MODEL_NAME")
 | RBT3 | hfl/rbt3 |
 | RBTL3 | hfl/rbtl3 |
 
-2. 使用PaddleHub
+### 使用PaddleHub
 
-依托[PaddleHub](https://github.com/PaddlePaddle/PaddleHub)，我们只需一行代码即可完成模型下载安装，十余行代码即可完成文本分类、序列标注、阅读理解等任务。
+依托[PaddleHub](https://github.com/PaddlePaddle/PaddleHub)，只需一行代码即可完成模型下载安装，十余行代码即可完成文本分类、序列标注、阅读理解等任务。
 
 ```
 import paddlehub as hub
@@ -155,12 +156,6 @@ module = hub.Module(name=MODULE_NAME)
 | RBTL3                 | [rbtl3](https://www.paddlepaddle.org.cn/hubdetail?name=rbtl3&en_category=SemanticModel) |
 
 
-### 测试任务数据
-我们提供部分任务数据，请查看`data`目录了解。
-压缩包内包含训练和测试数据，同一目录下的`README.md`标明数据来源。
-由于一部分数据需要原作者授权，故我们无法提供下载链接，敬请谅解。
-
-
 ## 模型对比
 针对大家比较关心的一些模型细节进行汇总如下。
 
@@ -392,7 +387,7 @@ A: 谷歌发布的中文BERT怎么用，这个就怎么用。
 A: 很遗憾，我不能提供相关代码，实现可以参考 [#10](https://github.com/ymcui/Chinese-BERT-wwm/issues/10) 和 [#13](https://github.com/ymcui/Chinese-BERT-wwm/issues/13)。
 
 **Q: 某某数据集在哪里下载？**  
-A: 请查看data目录。对于有版权的内容，请自行搜索或与原作者联系获取数据。
+A: 请查看`data`目录，任务目录下的`README.md`标明了数据来源。对于有版权的内容，请自行搜索或与原作者联系获取数据。
 
 **Q: 会有计划发布更大模型吗？比如BERT-large-wwm版本？**  
 A: 如果我们从实验中得到更好效果，会考虑发布更大的版本。

diff --git a/README_EN.md b/README_EN.md
@@ -11,7 +11,7 @@ For further accelerating Chinese natural language processing, we provide **Chine
 ## News
 2020/3/23 The models in this repository now can be easily accessed through [PaddleHub](https://github.com/PaddlePaddle/PaddleHub), check [Quick Load](#Quick-Load)
 
-**2020/2/26 We release a knowledge distillation toolkit [TextBrewer](https://github.com/airaria/TextBrewer)**
+2020/2/26 We release a knowledge distillation toolkit [TextBrewer](https://github.com/airaria/TextBrewer)
 
 2020/1/20 Happy Chinese New Year! We've released RBT3 and RBTL3 (3-layer RoBERTa-wwm-ext-base/large), check [Small Models](#Small-Models)
 
@@ -31,6 +31,7 @@ For further accelerating Chinese natural language processing, we provide **Chine
 |-|-|
 | [Introduction](#Introduction) | Introduction to BERT with Whole Word Masking (WWM) |
 | [Download](#Download) | Download links for Chinese BERT-wwm |
+| [Quick Load](#Quick-Load) | Learn how to quickly load our models through [🤗Transformers](https://github.com/huggingface/transformers) or [PaddleHub](https://github.com/PaddlePaddle/PaddleHub) |
 | [Model Comparison](#Model-Comparison) | Compare the models published in this repository |
 | [Baselines](#Baselines) | Baseline results for several Chinese NLP datasets (partial) |
 | [Small Models](#Small-Models) | 3-layer Transformer models |
@@ -98,8 +99,8 @@ chinese_wwm_L-12_H-768_A-12.zip
 `bert_config.json` and `vocab.txt` are identical to the original **`BERT-base, Chinese`** by Google。
 
 
-### Quick Load
-1. Huggingface-Transformers
+## Quick Load
+### Huggingface-Transformers
 
 With [Huggingface-Transformers](https://github.com/huggingface/transformers), the models above could be easily accessed and loaded through the following codes.
 ```
@@ -119,7 +120,7 @@ The actual model and its `MODEL_NAME` are listed below.
 | RBT3 | hfl/rbt3 |
 | RBTL3 | hfl/rbtl3 |
 
-2. PaddleHub
+### PaddleHub
 
 With [PaddleHub](https://github.com/PaddlePaddle/PaddleHub), we can download and install the model with one line of code.
 
@@ -130,7 +131,7 @@ module = hub.Module(name=MODULE_NAME)
 
 The actual model and its `MODULE_NAME` are listed below.
 
-| Original Model| MODULE_NAME |
+| Original Model | MODULE_NAME |
 | - | - |
 | RoBERTa-wwm-ext-large | [chinese-roberta-wwm-ext-large](https://www.paddlepaddle.org.cn/hubdetail?name=chinese-roberta-wwm-ext-large&en_category=SemanticModel) |
 | RoBERTa-wwm-ext       | [chinese-roberta-wwm-ext](https://www.paddlepaddle.org.cn/hubdetail?name=chinese-roberta-wwm-ext&en_category=SemanticModel) |
@@ -139,9 +140,6 @@ The actual model and its `MODULE_NAME` are listed below.
 | RBT3                  | [rbt3](https://www.paddlepaddle.org.cn/hubdetail?name=rbt3&en_category=SemanticModel) |
 | RBTL3                 | [rbtl3](https://www.paddlepaddle.org.cn/hubdetail?name=rbtl3&en_category=SemanticModel) |
 
-### Task Data
-We only provide the data that is publically available, check `data` directory.
-
 
 ## Model Comparison
 We list comparisons on the models that were released in this project.
@@ -361,7 +359,7 @@ A: Use it as if you are using original BERT. Note that, you don't need to do CWS
 A: Unfortunately, I am not be able to release the code at the moment. As implementation is quite easy, I would suggest you to read #10 and #13.
 
 **Q: How can I download XXXXX dataset?**  
-A: Please see `data` directory. For copyright reasons, some of the datasets are not publically available. In that case, please search on GitHub or consult original authors for accessing.
+A: We only provide the data that is publically available, check `data` directory. For copyright reasons, some of the datasets are not publically available. In that case, please search on GitHub or consult original authors for accessing.
 
 **Q: How to use this model?**  
 A: Use it as if you are using original BERT. Note that, you don't need to do CWS for your text, as wwm only change the pre-training input but not the input for down-stream tasks.