Update app.py

app.py

@@ -10,19 +10,27 @@ client = InferenceClient(
     api_key=HF_TOKEN,
 )
 
+# Modelos disponíveis
+MODELS = {
+    "LLaMA 70B": "meta-llama/Llama-3.3-70B-Instruct",
+    "Qwen 32B": "Qwen/QwQ-32B"
+}
+
 history_log = []
-show_history_flag = False 
+show_history_flag = False
+
+def chatbot_response(user_input, model_name):
+    model_id = MODELS[model_name]
 
-def chatbot_response(user_input):
     messages = [{"role": "user", "content": entry["Pergunta"]} for entry in history_log[-2:]]
     messages.append({"role": "user", "content": user_input})
     
     start_time = time.time()
     try:
         completion = client.chat.completions.create(
-            model="meta-llama/Llama-3.3-70B-Instruct", 
-            messages=messages, 
-            max_tokens=800,
+            model=model_id,
+            messages=messages,
+            max_tokens=8192 if "Qwen" in model_id else 800,
         )
         response = completion.choices[0].message['content']
     except Exception as e:
@@ -30,37 +38,48 @@ def chatbot_response(user_input):
     end_time = time.time()
     
     history_log.append({
+        "Modelo": model_name,
         "Pergunta": user_input,
         "Resposta": response,
         "Tempo de Resposta (s)": round(end_time - start_time, 2)
     })
-
-    print(history_log)
     
     return response
 
 with gr.Blocks(theme=gr.themes.Soft()) as demo:
-    gr.Markdown("# 🤖 Llama-70B Chatbot - SambaNova")
-    gr.Markdown("Modelo: meta-llama/Llama-3.3-70B-Instruct")
-    
-    chatbot = gr.Chatbot(height=500) 
-    msg = gr.Textbox(placeholder="Digite sua mensagem aqui...")
-    btn = gr.Button("Enviar", variant="primary") 
-    
-    def respond(message, chat_history):
-        response = chatbot_response(message)
+    with gr.Row():
+        with gr.Column(scale=1):
+            gr.Markdown("## ⚙️ Configurações")
+            model_selector = gr.Radio(
+                choices=list(MODELS.keys()),
+                label="Escolha o Modelo",
+                value="LLaMA 70B"
+            )
+            history_btn = gr.Button("Ver/Fechar Histórico", variant="secondary")
+            history_output = gr.JSON()
+        with gr.Column(scale=4):
+            gr.Markdown("# 🤖 Chatbot - SambaNova")
+            chatbot = gr.Chatbot(height=500)
+            msg = gr.Textbox(placeholder="Digite sua mensagem aqui...", show_label=False)
+            with gr.Row():
+                btn = gr.Button("Enviar", variant="primary")
+                history_btn = gr.Button("Ver/Fechar Histórico", variant="secondary")
+
+            history_output = gr.JSON()
+
+    def respond(message, chat_history, model_name):
+        response = chatbot_response(message, model_name)
         chat_history.append((message, response))
         return "", chat_history
-    
-    btn.click(respond, [msg, chatbot], [msg, chatbot])
-    
+
+    btn.click(respond, [msg, chatbot, model_selector], [msg, chatbot])
+    msg.submit(respond, [msg, chatbot, model_selector], [msg, chatbot])  # permite envio com Enter
+
     def toggle_history():
         global show_history_flag
         show_history_flag = not show_history_flag
         return history_log if show_history_flag else {}
-    
-    history_btn = gr.Button("Ver/Fechar Histórico", variant="secondary") 
-    history_output = gr.JSON()
+
     history_btn.click(toggle_history, inputs=[], outputs=history_output)
 
 if __name__ == "__main__":