update output

app.py

@@ -1,6 +1,8 @@
-import gradio as gr
 import os
+import gradio as gr
+import torch
 from transformers import pipeline
+from huggingface_hub import login
 
 # Lấy token từ biến môi trường
 token = os.getenv("TOKEN")
@@ -9,19 +11,41 @@ if token is None:
     raise ValueError("TOKEN environment variable not set")
 
 # Đăng nhập vào Hugging Face bằng token
-from huggingface_hub import login
 login(token=token)
 
 # Tải mô hình từ Hugging Face Hub
 model_name = "HuggingFaceH4/zephyr-orpo-141b-A35b-v0.1"
-classifier = pipeline("text-classification", model=model_name, token=token)
+pipe = pipeline(
+    "text-generation",
+    model=model_name,
+    device_map="auto",
+    torch_dtype=torch.bfloat16,
+)
 
 # Định nghĩa hàm xử lý
-def classify_text(text):
-    return classifier(text)
+def generate_text(user_input):
+    messages = [
+        {"role": "system", "content": "You are Zephyr, a helpful assistant."},
+        {"role": "user", "content": user_input},
+    ]
+    outputs = pipe(
+        messages,
+        max_new_tokens=512,
+        do_sample=True,
+        temperature=0.7,
+        top_k=50,
+        top_p=0.95,
+    )
+    return outputs[0]["generated_text"]
 
 # Tạo interface Gradio
-iface = gr.Interface(fn=classify_text, inputs="text", outputs="label")
+iface = gr.Interface(
+    fn=generate_text,
+    inputs="text",
+    outputs="text",
+    title="Zephyr - Text Generation",
+    description="Generate text responses using the Zephyr model."
+)
 
 # Khởi chạy ứng dụng
 iface.launch(share=True)