diff --git a/README.md b/README.md
index b9224b0..0d4cb54 100644
--- a/README.md
+++ b/README.md
@@ -1,5 +1,5 @@
 <div align="center">
-  <img src="./assets/page.jpg" alt="Logo" width="">
+  <img src="./assets/page.jpg" alt="Logo" width="500">
 </div>
 
 # SPEC-RL: Accelerating On-Policy Reinforcement Learning via Speculative Rollouts
@@ -15,9 +15,9 @@
   
 <!-- **Authors:** -->
 
-**_¹ [Bingshuai Liu](https://bingshuailiu.github.io),  [Ante Wang](), ¹ [Zijun Min](),_**
+**_¹ [Bingshuai Liu](https://bingshuailiu.github.io), ¹ Ante Wang, ¹ Zijun Min,_**
 
-**_² [Liang Yao](), ² [Haibo Zhang](), ² [Anxiang Zeng](), ¹ <sup>*</sup>[Jinsong Su]()_**
+**_² Liang Yao, ² Haibo Zhang, ² Anxiang Zeng, ¹ <sup>*</sup>Jinsong Su_**
 
 
 <!-- **Affiliations:** -->
diff --git a/requirements-npu.txt b/requirements-npu.txt
index 7d03869..63c17a7 100644
--- a/requirements-npu.txt
+++ b/requirements-npu.txt
@@ -18,4 +18,4 @@ mathruler
 torchdata
 einops
 qwen_vl_utils
-torchvision==0.20.1
+torchvision==0.23.0