Update app.py

app.py

@@ -6,7 +6,7 @@ from transformers import AutoTokenizer  # Import the tokenizer
 tokenizer = AutoTokenizer.from_pretrained("HuggingFaceH4/zephyr-7b-beta")
 client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
 
-# Define a maximum context length (tokens). Check your model's documentation!
+# Define a maximum context length (tokens).  Check your model's documentation!
 MAX_CONTEXT_LENGTH = 4096  # Example: Adjust this based on your model!
 
 nvc_prompt_template = r"""<|system|>
@@ -78,6 +78,7 @@ You are Roos, an NVC (Nonviolent Communication) Chatbot. Your goal is to help us
       - “Thank you for sharing with me. If you’d like to continue this conversation later, I’m here to help.”</s>
 """
 
+
 def count_tokens(text: str) -> int:
     """Counts the number of tokens in a given string."""
     return len(tokenizer.encode(text))
@@ -118,92 +119,53 @@ def respond(
     top_p,
 ):
     """Responds to a user message, maintaining conversation history, using special tokens and message list."""
-    # Allow a modifiable system prompt; if none is provided, fall back to the default.
-    formatted_system_message = system_message if system_message else nvc_prompt_template
 
-    truncated_history = truncate_history(
-        history,
-        formatted_system_message,
-        MAX_CONTEXT_LENGTH - max_tokens - 100  # Reserve space for the new message and some generation
-    )
+    formatted_system_message = nvc_prompt_template
+
+    truncated_history = truncate_history(history, formatted_system_message, MAX_CONTEXT_LENGTH - max_tokens - 100) # Reserve space for the new message and some generation
 
-    messages = [{"role": "system", "content": formatted_system_message}]  # Start with system message
+    messages = [{"role": "system", "content": formatted_system_message}] # Start with system message as before
     for user_msg, assistant_msg in truncated_history:
         if user_msg:
-            messages.append({"role": "user", "content": f"<|user|>\n{user_msg}</s>"})
+            messages.append({"role": "user", "content": f"<|user|>\n{user_msg}</s>"}) # Format history user message
         if assistant_msg:
-            messages.append({"role": "assistant", "content": f"<|assistant|>\n{assistant_msg}</s>"})
-    messages.append({"role": "user", "content": f"<|user|>\n{message}</s>"})
+            messages.append({"role": "assistant", "content": f"<|assistant|>\n{assistant_msg}</s>"}) # Format history assistant message
+
+    messages.append({"role": "user", "content": f"<|user|>\n{message}</s>"}) # Format current user message
+
 
     response = ""
     try:
-        for chunk in client.chat_completion(
-            messages,  # Send the messages list with the formatted content
-            max_tokens=max_tokens,
-            stream=True,
-            temperature=temperature,
-            top_p=top_p,
-        ):
-            token = chunk.choices[0].delta.content
-            response += token
-            yield response
+      for chunk in client.chat_completion(
+          messages, # Send the messages list again, but with formatted content
+          max_tokens=max_tokens,
+          stream=True,
+          temperature=temperature,
+          top_p=top_p,
+      ):
+          token = chunk.choices[0].delta.content
+          response += token
+          yield response
     except Exception as e:
-        print(f"An error occurred: {e}")
-        yield "I'm sorry, I encountered an error. Please try again."
-
-# --- Gradio Interface using Blocks ---
-with gr.Blocks() as demo:
-    # State to hold conversation history
-    state = gr.State([])
-
-    # Chatbot display
-    chatbot = gr.Chatbot(label="NVC Chatbot")
-
-    # System prompt textbox (modifiable)
-    with gr.Row():
-        system_prompt = gr.Textbox(value=nvc_prompt_template, label="System Prompt (modifiable)")
-
-    # Controls for generation parameters
-    with gr.Row():
-        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")
-        top_p_slider = gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)")
-
-    # Button to clear conversation memory
-    with gr.Row():
-        clear_button = gr.Button("Clear Memory")
-
-    # User input area and send button
-    with gr.Row():
-        user_input = gr.Textbox(label="Your Message", placeholder="Type your message here...")
-        send_button = gr.Button("Send")
-
-    def chat_step(message, history, system_message, max_tokens, temperature, top_p):
-        # Call the respond generator and accumulate the final response.
-        gen = respond(message, history, system_message, max_tokens, temperature, top_p)
-        response = ""
-        for r in gen:
-            response = r  # In a streaming scenario, you might update the UI incrementally.
-        history.append((message, response))
-        return history, history
-
-    # Trigger the chat step on button click or when submitting the textbox.
-    send_button.click(
-        chat_step,
-        inputs=[user_input, state, system_prompt, max_tokens_slider, temperature_slider, top_p_slider],
-        outputs=[chatbot, state],
-    )
-    user_input.submit(
-        chat_step,
-        inputs=[user_input, state, system_prompt, max_tokens_slider, temperature_slider, top_p_slider],
-        outputs=[chatbot, state],
-    )
-
-    # Clear memory: resets both the chatbot display and the state.
-    def clear_history():
-        return [], []
-
-    clear_button.click(clear_history, inputs=[], outputs=[chatbot, state])
+      print(f"An error occurred: {e}") # It's a good practice add a try-except block
+      yield "I'm sorry, I encountered an error. Please try again."
+
+# --- Gradio Interface ---
+demo = gr.ChatInterface(
+    respond,
+    additional_inputs=[
+        gr.Textbox(value=nvc_prompt_template, label="System message", visible=False), # Set the NVC prompt as default and hide the system message box
+        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
+        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
+        gr.Slider(
+            minimum=0.1,
+            maximum=1.0,
+            value=0.95,
+            step=0.05,
+            label="Top-p (nucleus sampling)",
+        ),
+    ],
+)
 
 if __name__ == "__main__":
     demo.launch()