Update app.py

app.py

@@ -2,94 +2,29 @@ import torch
 from transformers import AutoModelForCausalLM, AutoTokenizer
 import gradio as gr
 
-# 1. 모델 초기화
+# 모델 초기화 (간소화 버전)
 device = "cuda" if torch.cuda.is_available() else "cpu"
-print(f"Using device: {device}")
+model = AutoModelForCausalLM.from_pretrained("naver-hyperclovax/HyperCLOVAX-SEED-Text-Instruct-0.5B").to(device)
+tokenizer = AutoTokenizer.from_pretrained("naver-hyperclovax/HyperCLOVAX-SEED-Text-Instruct-0.5B")
 
-try:
-    model = AutoModelForCausalLM.from_pretrained(
-        "naver-hyperclovax/HyperCLOVAX-SEED-Text-Instruct-0.5B",
-        torch_dtype=torch.float16 if device == "cuda" else torch.float32
-    ).to(device)
-    
-    tokenizer = AutoTokenizer.from_pretrained(
-        "naver-hyperclovax/HyperCLOVAX-SEED-Text-Instruct-0.5B"
-    )
-    print("Model loaded successfully")
-except Exception as e:
-    print(f"Model loading failed: {e}")
-    raise
-
-# 2. 채팅 생성 함수
-def generate_response(chat_history, user_input):
-    # 대화 기록 업데이트
-    chat_history.append({"role": "user", "content": user_input})
-    
-    # 템플릿 적용
-    inputs = tokenizer.apply_chat_template(
-        chat_history,
-        add_generation_prompt=True,
-        return_tensors="pt"
-    ).to(device)
-    
-    # 응답 생성
-    output_ids = model.generate(
-        inputs,
-        max_length=1024,
-        temperature=0.7,
-        top_p=0.9,
-        do_sample=True,
-        eos_token_id=tokenizer.eos_token_id
-    )
-    
-    # 응답 디코딩
-    response = tokenizer.decode(
-        output_ids[0][inputs.shape[1]:], 
-        skip_special_tokens=True
-    )
-    
-    chat_history.append({"role": "assistant", "content": response})
-    return response
+def respond(message, history):
+    # 대화 기록을 모델 입력 형식으로 변환
+    chat = [
+        {"role": "system", "content": "당신은 네이버의 CLOVA X AI입니다."},
+        *[{"role": "user" if h[0] == message else "assistant", "content": h[1]} for h in history],
+        {"role": "user", "content": message}
+    ]
+    
+    inputs = tokenizer.apply_chat_template(chat, return_tensors="pt").to(device)
+    outputs = model.generate(inputs, max_length=1024, temperature=0.7)
+    return tokenizer.decode(outputs[0][inputs.shape[1]:], skip_special_tokens=True)
 
-# 3. Gradio 인터페이스
-def chat_interface(user_input, chat_history_ui):
-    # 초기 시스템 메시지
-    if not chat_history_ui:
-        chat_history = [
-            {"role": "system", "content": "당신은 네이버의 CLOVA X AI입니다."}
-        ]
-    else:
-        chat_history = []
-        for msg in chat_history_ui:
-            chat_history.extend([
-                {"role": "user", "content": msg[0]},
-                {"role": "assistant", "content": msg[1]}
-            ])
-    
-    # 응답 생성
-    bot_response = generate_response(chat_history, user_input)
-    chat_history_ui.append((user_input, bot_response))
-    
-    return "", chat_history_ui
-
-# 4. 앱 실행
-with gr.Blocks(theme=gr.themes.Soft()) as demo:
-    gr.Markdown("## 🧑💻 HyperCLOVAX-SEED 챗봇")
-    
-    chatbot = gr.Chatbot(height=500)
-    msg = gr.Textbox(label="메시지 입력")
-    clear = gr.Button("초기화")
-    
-    msg.submit(
-        chat_interface,
-        [msg, chatbot],
-        [msg, chatbot]
-    )
-    clear.click(lambda: None, None, chatbot, queue=False)
+# Gradio 인터페이스 (최소화)
+demo = gr.ChatInterface(
+    respond,
+    title="CLOVA X 챗봇",
+    description="네이버 HyperCLOVAX-SEED 기반 챗봇"
+)
 
 if __name__ == "__main__":
-    demo.launch(
-        server_name="0.0.0.0",
-        server_port=7860,
-        share=False
-    )
+    demo.launch(server_name="0.0.0.0", server_port=7860)