Update app.py

app.py

@@ -202,14 +202,15 @@ def generate_video(model_name: str, text: str, video_path: str,
     buffer = ""
     for new_text in streamer:
         buffer += new_text
+        buffer = buffer.replace("<|im_end|>", "")
         time.sleep(0.01)
         yield buffer
 
 # Define examples for image and video inference
 image_examples = [
-    ["OCR the table", "example/image1.png"],
-    ["Explain the scene", "example/image2.jpg"],
-    ["Fill the correct numbers", "example/image3.png"],
+    ["ocr the image", "example/image1.png"],
+    ["explain the scene", "example/image2.jpg"],
+    ["fill the correct numbers", "example/image3.png"],
 ]
 
 video_examples = [