Update app.py

app.py

@@ -1,7 +1,7 @@
 import gradio as gr
 import requests
-import os
 import json
+import os
 
 # API and environment variables
 API_KEY = os.getenv('API_KEY')
@@ -16,18 +16,22 @@ headers = {
 # Base system message
 BASE_SYSTEM_MESSAGE = "I carefully provide accurate, factual, thoughtful, nuanced answers and am brilliant at reasoning."
 
-def call_nvidia_api(history, system_message, max_tokens, temperature, top_p):
-    """Calls the NVIDIA API to generate a response."""
-    messages = [{"role": "system", "content": system_message}]
-    messages.extend([{"role": "user", "content": msg} for msg, _ in history])
+def clear_chat():
+    """Clears the chat history and message state."""
+    print("Clearing chat...")
+    chat_history_state.value = []
+    chatbot.textbox.value = ""
 
+def call_nvidia_api(api_history, max_tokens, temperature, top_p):
+    """Calls the NVIDIA API to generate a response."""
     payload = {
-        "messages": messages,
+        "messages": api_history,
         "temperature": temperature,
         "top_p": top_p,
         "max_tokens": max_tokens,
         "stream": False
     }
+    print(f"Payload enviado: {payload}")
     session = requests.Session()
     response = session.post(INVOKE_URL, headers=headers, json=payload)
     while response.status_code == 202:
@@ -36,31 +40,37 @@ def call_nvidia_api(history, system_message, max_tokens, temperature, top_p):
         response = session.get(fetch_url, headers=headers)
         response.raise_for_status()
     response_body = response.json()
+    print(f"Payload recebido: {response_body}")
     if response_body.get("choices"):
         assistant_message = response_body["choices"][0]["message"]["content"]
         return assistant_message
     else:
-        return "Sorry, there was an error generating the response."
+        return "Desculpe, ocorreu um erro ao gerar a resposta."
 
-def chatbot_submit(message, chat_history, system_message, max_tokens_val, temperature_val, top_p_val):
-    """Submits the user message to the chatbot and updates the chat history."""
-    chat_history.append([message, ""])  # Add user message to history
+def chatbot_submit(message, chat_history, api_history, system_message, max_tokens_val, temperature_val, top_p_val):
+    """Submits the user message and updates both histories."""
+    # Update Gradio history
+    chat_history.append([message, ""])
 
-    # Call NVIDIA API to generate a response
-    assistant_message = call_nvidia_api(chat_history, system_message, max_tokens_val, temperature_val, top_p_val)
+    # Update API history
+    api_history.append({"role": "user", "content": message})
 
-    # Update history with assistant's response
+    # Call NVIDIA API
+    assistant_message = call_nvidia_api(api_history, max_tokens_val, temperature_val, top_p_val)
+
+    # Update Gradio history with response
     chat_history[-1][1] = assistant_message
 
-    return assistant_message, chat_history
+    # Update API history with response
+    api_history.append({"role": "assistant", "content": assistant_message})
+
+    return assistant_message, chat_history, api_history
 
-# Gradio interface setup
 chat_history_state = gr.State([])
 system_msg = gr.Textbox(BASE_SYSTEM_MESSAGE, label="System Message", placeholder="System prompt.", lines=5)
 max_tokens = gr.Slider(20, 1024, label="Max Tokens", step=20, value=1024)
 temperature = gr.Slider(0.0, 1.0, label="Temperature", step=0.1, value=0.2)
 top_p = gr.Slider(0.0, 1.0, label="Top P", step=0.05, value=0.7)
-
 with gr.Blocks() as demo:
     chat_history_state = gr.State([])
     chatbot = gr.ChatInterface(