Update app.py

app.py

@@ -28,6 +28,7 @@ model = LlavaForConditionalGeneration.from_pretrained(
 )
 
 model.to("cuda:0")
+model.generation_config.eos_token_id = tokenizer.convert_tokens_to_ids("<|eot_id|>")
 
 @spaces.GPU
 def bot_streaming(message, history):
@@ -68,8 +69,9 @@ def bot_streaming(message, history):
     yield generated_text_without_prompt
 
 
-demo = gr.ChatInterface(fn=bot_streaming, css=CSS, title="LLaVA Llama-3-8B", examples=[{"text": "What is on the flower?", "files":["./bee.jpg"]},
+with gr.Blocks(css=CSS) as demo:
+    chatbot = gr.ChatInterface(fn=bot_streaming, title="LLaVA Llama-3-8B", examples=[{"text": "What is on the flower?", "files":["./bee.jpg"]},
                                                                       {"text": "How to make this pastry?", "files":["./baklava.png"]}], 
-                        description="Try [LLaVA Llama-3-8B](https://huggingface.co/xtuner/llava-llama-3-8b-v1_1-transformers). Upload an image and start chatting about it, or simply try one of the examples below. If you don't upload an image, you will receive an error.",
-                        stop_btn="Stop Generation", multimodal=True)
+                            description="Try [LLaVA Llama-3-8B](https://huggingface.co/xtuner/llava-llama-3-8b-v1_1-transformers). Upload an image and start chatting about it, or simply try one of the examples below. If you don't upload an image, you will receive an error.",
+                            stop_btn="Stop Generation", multimodal=True)
 demo.launch(debug=True)