diff --git a/README.md b/README.md
index 0035424..a0df028 100644
--- a/README.md
+++ b/README.md
@@ -47,5 +47,10 @@ To reproduce the results of TIGER or train TIGER, you are supposed to download a
 └── data_process.py
 
 ```
+Then, you can train TIGER with the following command:
+'''
+python main.py -dataset drugbank/kegg/ogbl-biokg -extractor khop-subtree/randomWalk/probability
+'''
 
+The hyper-parameters used to train TIGER on above three datasets are shown in our paper.