doc: little update

app.py

@@ -72,9 +72,9 @@ MARKDOWN = \
        
     </div>
 
-    Frame In-N-Out expands the first frame condition to a broader canvas region by setting top left and bottom right expansion amount, 
-    and users could provide motion trajectory to existing objects or provide breaking new identity to enter the scene with motion trajectory, or both. <br>
-    The model we used here is <b>Wan2.2-5B</b> trained on our Frame In-N-Out control mechanism.
+    Frame In-N-Out extends the first-frame conditioning to a larger spatial canvas by specifying top-left and bottom-right expansion offsets. 
+    Further, it allows users to assign motion trajectories to existing objects, introduce new identities that enter the scene with their own trajectories, or both.<br>
+    The model we used here is [<b>Wan2.2-5B</b> V1.6](https://huggingface.co/uva-cv-lab/FrameINO_Wan2.2_5B_Stage2_MotionINO_v1.6) trained on our Frame In-N-Out control mechanism.
 
 
     <br>