Update README.md

README.md

@@ -69,6 +69,11 @@ We trained models using datasets of different sizes (200,000, 600,000, and 1,000
 |LR_scheduler | linear |
 
 ## Use model
+Please note that the input should be formatted as follows in both **training** and **inference**.
+``` python
+Human: {input} \n\nAssistant:
+```
+
 BELLE can be easily loaded with AutoModelForCausalLM.
 ``` python
 from transformers import AutoTokenizer, AutoModelForCausalLM
@@ -142,8 +147,12 @@ BELLE模型以Bloomz-7b1-mt为基础，在 0.2M 条中文数据上，结合Stanf
 |LR_scheduler | linear |
 
 ## 使用模型
-通过AutoModelForCausalLM即可直接载入模型并使用。
+请注意，在训练和推理时，模型的输入应该处理成如下形式：
+``` python
+Human: {input} \n\nAssistant:
+```
 
+通过AutoModelForCausalLM即可直接载入模型并使用。
 ``` python
 from transformers import AutoTokenizer, AutoModelForCausalLM
 import sys