Update app.py

app.py

@@ -8,7 +8,7 @@ model_file_path = hf_hub_download(
     filename="llama-2-7b.Q4_0.gguf"
 )
 
-# Try to load the model from the downloaded file
+# Initialize the Llama model
 try:
     llm_llama_cpp = Llama(
         model_path=model_file_path,  # Path where the model is downloaded
@@ -16,26 +16,28 @@ try:
         n_ctx=4096  # Set context window to match model's full capacity
     )
 
-    # Function that generates a response using the Llama model
+    # Define the function for generating text with streaming
     def talk(prompt, history):
         try:
-            # Generate text with streaming
             response_stream = llm_llama_cpp.create_completion(
                 prompt,
                 max_tokens=200,  # You can adjust the max tokens as needed
                 stream=True
             )
 
+            # Prepare a response variable to store the final result
             response = ""
             for chunk in response_stream:
-                # Extract the text from the current chunk
+                # Extract and accumulate the text from each chunk
                 if 'choices' in chunk and len(chunk['choices']) > 0 and 'text' in chunk['choices'][0]:
                     response += chunk['choices'][0]['text']
+                    print(f"Streaming: {chunk['choices'][0]['text']}", end="", flush=True)
 
-            # Return the full response after streaming is done
+            # After the stream is complete, return the final response
             return response
 
         except Exception as e:
+            print(f"Error in generating response: {e}")
             return f"Error with llama-cpp-python: {e}"
 
 except FileNotFoundError:
@@ -47,7 +49,7 @@ except Exception as e:
 TITLE = "AI Copilot for Diabetes Patients"
 DESCRIPTION = "I provide answers to concerns related to Diabetes"
 
-# Design chatbot interface (fixed `likeable` argument)
+# Design chatbot interface (fixed `likeable` argument, deprecated params removed)
 demo = gr.ChatInterface(
     fn=talk,  # The function that processes user input and returns the response
     chatbot=gr.Chatbot(
@@ -55,7 +57,7 @@ demo = gr.ChatInterface(
         show_share_button=True,
         show_copy_button=True,
         layout="bubble",  # Display messages in bubble format
-        bubble_full_width=False,
+        type="messages",  # Use OpenAI-style message format
     ),
     theme="Soft",  # Soft theme for the UI
     examples=[["what is Diabetes?"]],  # Example query to get started
@@ -64,4 +66,4 @@ demo = gr.ChatInterface(
 )
 
 # Launch the chatbot interface
-demo.launch()
+demo.launch()