Update README.md

README.md

@@ -16,7 +16,9 @@ Please follow the license of the original model. This model could **NOT** run on
 
 ## How To Use
 
-INT4 Inference on CUDA(4X80G), please note int2 **may be slower** than to int4 on CUDA due to kernel issue.
+### INT2 Inference on CUDA(4X80G)
+
+please note int2 **may be slower** than int4 on CUDA due to kernel issue.
 
 **To prevent potential overflow, we recommend using the CPU version.**
 
@@ -238,7 +240,7 @@ But wait, maybe the question is more literal. If the hunter successfully shoots
 
 ~~~
 
-### INT4 Inference on CPU
+### INT2 Inference on CPU
 
 Requirements