Update app.py

app.py

@@ -6,7 +6,7 @@ import gradio as gr
 # Use GPU if available
 device = "cuda" if torch.cuda.is_available() else "cpu"
 
-# Base model and adapter paths (updated for Hugging Face repo)
+# Base model and adapter paths
 base_model_name = "microsoft/phi-2"  # Pull from HF Hub directly
 adapter_path = "Shriti09/Microsoft-Phi-QLora"  # Update with your Hugging Face repo path
 
@@ -30,7 +30,7 @@ print("✅ Model ready for inference!")
 
 # Chat function with history
 def chat_fn(message, history):
-    # Combine conversation history into one prompt
+    # Convert history to the required format for gr.Chatbot (list of dictionaries with role and content)
     full_prompt = ""
     for user_msg, bot_msg in history:
         full_prompt += f"User: {user_msg}\nAI: {bot_msg}\n"
@@ -53,8 +53,10 @@ def chat_fn(message, history):
     response = tokenizer.decode(outputs[0], skip_special_tokens=True)
     response = response.split("AI:")[-1].strip()
 
-    # Append to history
-    history.append((message, response))
+    # Append to history in the correct format for gr.Chatbot (list of dictionaries)
+    history.append({"role": "user", "content": message})
+    history.append({"role": "assistant", "content": response})
+
     return history, history
 
 # Gradio UI
@@ -72,5 +74,5 @@ with gr.Blocks(theme=gr.themes.Soft()) as demo:
     clear.click(lambda: [], None, chatbot)
     clear.click(lambda: [], None, state)
 
-# Run the app without the 'concurrency_count' argument
-demo.queue().launch()
+# Run the app without the 'concurrency_count' argument and share the app publicly
+demo.queue().launch(share=True)