Refactor call_nvidia_api and chat functions

app.py

@@ -26,16 +26,14 @@ def user(message, history):
     history = history or []
     history.append({"role": "user", "content": message})
     return history
-def call_nvidia_api(history):
-    # Preparar o payload com o histórico de chat formatado
+def call_nvidia_api(history, max_tokens, temperature, top_p):
     payload = {
         "messages": history,
-        "temperature": 0.7,
-        "top_p": 0.95,
-        "max_tokens": 500,
-        "seed": 42,
-        "stream": False 
-    } 
+        "temperature": temperature,
+        "top_p": top_p,
+        "max_tokens": max_tokens,
+        "stream": False
+    }
 
     session = requests.Session()
     response = session.post(INVOKE_URL, headers=headers, json=payload)
@@ -55,11 +53,11 @@ def call_nvidia_api(history):
     
     return history
 
-def chat(history, system_message):
+def chat(history, system_message, max_tokens, temperature, top_p, top_k, repetition_penalty):
     print("Starting chat...")
-    # Atualizar o histórico com a resposta do assistente
-    updated_history = call_nvidia_api(history)
+    updated_history = call_nvidia_api(history, max_tokens, temperature, top_p)
     return updated_history, ""
+
 # Gradio interface setup
 with gr.Blocks() as demo:
     with gr.Row():