Provide an easy call for predicting description.

incomplete_utterance_rewriting/README.md

@@ -21,6 +21,7 @@ If you find our code useful for you, please consider citing our paper:
 - [Download and Preprocess Dataset](#data)
 - [Train Model](#train)
 - [Evaluate Model](#evaluate)
+- [Predict Model](#predict)
 - [Pre-trained Models](#pre-trained-models)
 
 ## Requirement
@@ -135,6 +136,10 @@ Next, we will provide all pre-trained models to reproduce results reported in ou
 python evaluate.py --model_file ../pretrianed_weights/rewrite.tar.gz --test_file ../dataset/Multi/test.txt
 ```
 
+## Predict
+
+We provide an easy function call to predict a rewriting sentence given a specific dialogue context in `src/predict.py`. You could follow it to customize your self.
+
 ## Pre-trained Models
 
 

incomplete_utterance_rewriting/README_zh.md

@@ -21,6 +21,7 @@
 - [数据集下载与预处理](#数据集下载与预处理)
 - [训练模型](#训练模型)
 - [评测模型](#评测模型)
+- [预测模型](#预测模型)
 - [预训练模型权重](#预训练模型权重)
 
 ## 依赖安装
@@ -133,6 +134,10 @@ python evaluate.py --model_file model.tar.gz --test_file ../dataset/Multi/test.t
 python evaluate.py --model_file ../pretrianed_weights/rewrite.tar.gz --test_file ../dataset/Multi/test.txt
 ```
 
+## 预测模型
+
+我们在`src/predict.py`中提供了一个简单的函数调用，来实现在python code中为一个给定的对话上下文生成改写句。读者可根据实际项目需要进行改造。
+
 
 ## 预训练模型权重