Update app.py

app.py

@@ -23,18 +23,32 @@ def load_model():
             low_cpu_mem_usage=True
         )
 
-def chat(prompt):
+def respond(message, history):
     load_model()
-    inputs = tokenizer(prompt, return_tensors="pt")
+
+    # Ghép lịch sử chat
+    prompt = ""
+    for user, bot in history:
+        prompt += f"User: {user}\nAssistant: {bot}\n"
+    prompt += f"User: {message}\nAssistant:"
+
+    inputs = tokenizer(prompt, return_tensors="pt", truncation=True)
     outputs = model.generate(
         **inputs,
-        max_new_tokens=128
+        max_new_tokens=150,
+        do_sample=True,
+        temperature=0.7,
+        top_p=0.9
     )
-    return tokenizer.decode(outputs[0], skip_special_tokens=True)
 
-gr.Interface(
-    fn=chat,
-    inputs="textbox",
-    outputs="textbox",
-    title="Gemma3 270M Cloud"
+    text = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    reply = text.split("Assistant:")[-1].strip()
+
+    return reply
+
+gr.ChatInterface(
+    fn=respond,
+    title="🤖 Gemma3 270M Cloud Chat",
+    description="Gemma3 270M chạy cloud miễn phí trên Hugging Face Spaces",
+    theme="dark"
 ).launch(server_name="0.0.0.0")