Update app.py

app.py

@@ -330,8 +330,7 @@ with gr.Blocks(css=css, theme="bethecloud/storj_theme") as demo:
             model_choice = gr.Radio(choices=[
                  "Camel-Doc-OCR-080125(v2)", "OCRFlux-3B",
                  "ViGoRL-MCTS-SFT-3B", "Behemoth-3B-070225",
-                 "MonkeyOCR-pro-1.2B"
-            ],
+                 "MonkeyOCR-pro-1.2B"],
                                     label="Select Model",
                                     value="Camel-Doc-OCR-080125(v2)")
             
@@ -339,7 +338,8 @@ with gr.Blocks(css=css, theme="bethecloud/storj_theme") as demo:
             gr.Markdown("> Camel-Doc-OCR-080125 is a specialized vision-language model, fine-tuned from Qwen2.5-VL-7B-Instruct, and excels at document retrieval, content extraction, and analysis recognition for both structured and unstructured digital documents. OCRFlux-3B is a 3B-parameter vision-language model optimized for high-quality OCR on PDFs and images, excelling in converting documents to clean Markdown text and supporting features like cross-page table/paragraph merging.")    
             gr.Markdown("> Both ViGoRL-MCTS-SFT-3b-Spatial and 7b-Spatial are vision-language models that use multi-turn visually grounded reinforcement learning for precise spatial reasoning and visual grounding, with the 3b and 7b variants differing mainly in their architectural size for fine-grained visual tasks.")
             gr.Markdown("> Behemoth-3B-070225-post0.1 is an advanced 3B parameter model tailored for extensive multimodal comprehension, document parsing, and possibly generalized OCR/vision-language tasks. MonkeyOCR-pro-1.2B is a lightweight OCR model focusing on high-accuracy text extraction from images and scanned documents, suitable for resource-constrained environments.")    
-  
+            gr.Markdown("> ⚠️ Note: Models in this space may not perform well on video inference tasks.")
+            
     # Define the submit button actions
     image_submit.click(fn=generate_image,
                        inputs=[