Update app.py

app.py

@@ -20,31 +20,30 @@ For more information on Hugging Face Inference API support, please check:
 https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
 """
 
-# Initialize the Hugging Face model client (make sure you have access)
+# Initialize your Hugging Face model client.
 client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
 
-def respond(message, history: list[tuple[str, str]], system_message, max_tokens, temperature, top_p, file_content):
+def respond(message, history: list[dict], system_message, max_tokens, temperature, top_p, file_content):
     """
     Calls the model (in non-streaming mode) to get a complete response.
-    If file_content is non-empty, it is appended to the system message (context).
+    The file content is appended to the system message as context.
+    Expects conversation history in the format: 
+      [{"role": "user", "content": ...}, {"role": "assistant", "content": ...}, ...]
     """
     if file_content and file_content.strip():
         system_message += "\n\nFile content:\n" + file_content
 
-    # Build messages list in the expected format.
+    # Build messages list for the API request.
     messages = [{"role": "system", "content": system_message}]
-    for user_msg, assistant_msg in history:
-        if user_msg:
-            messages.append({"role": "user", "content": user_msg})
-        if assistant_msg:
-            messages.append({"role": "assistant", "content": assistant_msg})
+    for entry in history:
+        messages.append(entry)
     messages.append({"role": "user", "content": message})
     
     try:
         completion = client.chat_completion(
             messages,
             max_tokens=max_tokens,
-            stream=False,  # Non-streaming mode for simplicity
+            stream=False,  # Non-streaming mode for simplicity.
             temperature=temperature,
             top_p=top_p,
         )
@@ -54,7 +53,7 @@ def respond(message, history: list[tuple[str, str]], system_message, max_tokens,
     return response
 
 ###############################################################################
-#                        File Upload & Parsing Functions                        #
+#                       File Upload & Parsing Functions                        #
 ###############################################################################
 
 def parse_file(file_obj):
@@ -98,31 +97,31 @@ def load_files(files):
 ###############################################################################
 
 with gr.Blocks() as demo:
-    gr.Markdown("# Combined Chat & File Upload App")
+    gr.Markdown("# **Combined Chat & File Upload App**")
     gr.Markdown(
         """
-This app allows you to upload file(s) (PDF or TXT) to provide context for the AI.
-Once files are uploaded, their contents are automatically parsed and used in every conversation.
-Simply upload a file and then start chatting.
+This app allows you to upload file(s) (e.g., PDF or TXT) and chat with an AI assistant that uses the uploaded file(s) for context throughout the conversation.
+
+- **Upload File(s):** The file contents are automatically parsed and stored.
+- **Chat:** Your message, along with the uploaded file content, is sent to the AI on every prompt.
         """
     )
     
-    # State to hold file content (the concatenated text of uploaded files)
+    # State to hold the concatenated file content and conversation history.
     file_content_state = gr.State("")
-    # State to hold the conversation history (list of (user, assistant) tuples)
-    chat_history_state = gr.State([])
-
-    # --- File Upload ---
-    # Using type="filepath" so that we get a file path that can be opened later.
+    chat_history_state = gr.State([])  # List of dictionaries in the form {"role": "user"/"assistant", "content": ...}
+    
+    # --- File Upload Section ---
+    # Use type="filepath" so that we get file paths for processing.
     file_input = gr.File(label="Upload File(s)", file_count="multiple", type="filepath")
-    # Automatically process files when they are uploaded.
+    # Automatically process files upon upload.
     file_input.change(fn=load_files, inputs=file_input, outputs=file_content_state)
     
     gr.Markdown("## Chat")
-    chatbot = gr.Chatbot(label="Chat History")
+    chatbot = gr.Chatbot(label="Chat History", type="messages")
     user_input = gr.Textbox(label="Your Message", placeholder="Type your message here...", lines=2)
     
-    # Additional model parameters:
+    # Additional model parameters (adjustable)
     system_prompt = gr.Textbox(label="System Message", value="You are a helpful AI assistant.", interactive=True)
     max_tokens_slider = gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max New Tokens")
     temperature_slider = gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature")
@@ -131,15 +130,15 @@ Simply upload a file and then start chatting.
     def chat_fn(user_msg, history, file_content, system_msg, max_tokens, temperature, top_p):
         if not user_msg.strip():
             return "", history
-        # Append the user's message to the conversation history.
-        history.append((user_msg, ""))
-        # Call the respond function (non-streaming) to get a complete answer.
+        # Append user's message (in the required format).
+        history.append({"role": "user", "content": user_msg})
+        # Get the AI's response.
         response = respond(user_msg, history, system_msg, max_tokens, temperature, top_p, file_content)
-        # Update the last entry in the conversation with the response.
-        history[-1] = (user_msg, response)
+        # Append the assistant's response.
+        history.append({"role": "assistant", "content": response})
         return "", history
     
-    # When user submits a message, call chat_fn.
+    # Trigger sending message on Enter in the textbox.
     user_input.submit(
         fn=chat_fn,
         inputs=[user_input, chat_history_state, file_content_state, system_prompt, max_tokens_slider, temperature_slider, top_p_slider],
@@ -147,7 +146,16 @@ Simply upload a file and then start chatting.
         queue=True
     )
     
-    demo.launch(server_name="0.0.0.0", server_port=7860)
+    # Also add a "Send" button.
+    send_button = gr.Button("Send")
+    send_button.click(
+        fn=chat_fn,
+        inputs=[user_input, chat_history_state, file_content_state, system_prompt, max_tokens_slider, temperature_slider, top_p_slider],
+        outputs=[user_input, chatbot],
+        queue=True
+    )
+    
+    demo.launch(server_name="0.0.0.0", server_port=7860, share=True)
 
 if __name__ == "__main__":
     demo.launch()