Update app.py

app.py

@@ -13,6 +13,14 @@ import re
 
 DESCRIPTION = "# LLaVA 🌋"
 
+model_id = "llava-hf/llava-1.5-7b-hf"
+quantization_config = BitsAndBytesConfig(
+    load_in_4bit=True,
+    bnb_4bit_compute_dtype=torch.float16
+)
+pipe = pipeline("image-to-text", model=model_id, model_kwargs={"quantization_config": quantization_config})
+
+
 def extract_response_pairs(text):
     pattern = re.compile(r'(USER:.*?)ASSISTANT:(.*?)(?:$|USER:)', re.DOTALL)
     matches = pattern.findall(text)
@@ -61,11 +69,15 @@ css = """
   """
 with gr.Blocks(css="style.css") as demo:
     gr.Markdown(DESCRIPTION)
-    gr.Markdown("## LLaVA, one of the greatest multimodal chat models is now available in transformers with 4-bit quantization!")
+    gr.Markdown("**LLaVA, one of the greatest multimodal chat models is now available in transformers with 4-bit quantization! ⚡️  **")
+    gr.Markdown("**Try it in this demo 🤗 **")
+
     chatbot = gr.Chatbot(label="Chat", show_label=False)
+    gr.Markdown("Input image and text and start chatting 👇")
     with gr.Row():
+      
       image = gr.Image(type="pil")
-      text_input = gr.Text(label="Chat Input", show_label=False, max_lines=1, container=False)
+      text_input = gr.Text(label="Chat Input", show_label=False, max_lines=3, container=False)