Update app.py

app.py

@@ -1,64 +1,374 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+import torch
 
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
+try:
+    from unsloth import FastLanguageModel
+except ImportError:
+    print("Unsloth가 설치되어 있지 않습니다. 설치 중...")
+    import subprocess
+    subprocess.check_call(["pip", "install", "unsloth"])
+    from unsloth import FastLanguageModel
 
+# Hugging Face에 업로드된 모델 사용
+MODEL_NAME = "huggingface-KREW/Llama-3.1-8B-Spider-SQL-Ko"
 
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
+print(f"Loading model from Hugging Face: {MODEL_NAME}")
 
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
+# Unsloth를 사용하여 모델과 토크나이저 로드
+try:
+    model, tokenizer = FastLanguageModel.from_pretrained(
+        model_name=MODEL_NAME,
+        max_seq_length=2048,
+        dtype=None,  # 자동 감지
+        load_in_4bit=True,  # 4비트 양자화 사용
+    )
+    
+    # 추론 모드로 설정
+    FastLanguageModel.for_inference(model)
+    print("Model loaded successfully with Unsloth!")
+    
+except Exception as e:
+    print(f"Error loading model with Unsloth: {e}")
+    print("\n모델이 Hugging Face에 제대로 업로드되지 않았을 수 있습니다.")
+    print("로컬 모델을 사용하거나 모델을 다시 업로드해주세요.")
+    raise
 
-    messages.append({"role": "user", "content": message})
+# Example databases and questions
+examples = [
+    {
+        "db_id": "department_management",
+        "question": "각 부서별 직원 수를 보여주세요.",
+        "schema": """데이터베이스 스키마:
+테이블: department
+컬럼:
+  - Department_ID (number) (기본 키)
+  - Name (text)
+  - Creation (text)
+  - Ranking (number)
+  - Budget_in_Billions (number)
+  - Num_Employees (number)
+테이블: head
+컬럼:
+  - head_ID (number) (기본 키)
+  - name (text)
+  - born_state (text)
+  - age (number)
+테이블: management
+컬럼:
+  - department_ID (number) (기본 키)
+  - head_ID (number)
+  - temporary_acting (text)
 
-    response = ""
+외래 키 관계:
+  - management.head_ID → head.head_ID
+  - management.department_ID → department.Department_ID"""
+    },
+    {
+        "db_id": "concert_singer",
+        "question": "가장 많은 콘서트를 연 가수는 누구인가요?",
+        "schema": """데이터베이스 스키마:
+테이블: singer
+컬럼:
+  - Singer_ID (number) (기본 키)
+  - Name (text)
+  - Country (text)
+  - Song_Name (text)
+  - Song_release_year (text)
+  - Age (number)
+  - Is_male (text)
+테이블: concert
+컬럼:
+  - concert_ID (number) (기본 키)
+  - concert_Name (text)
+  - Theme (text)
+  - Stadium_ID (number)
+  - Year (text)
+테이블: singer_in_concert
+컬럼:
+  - concert_ID (number)
+  - Singer_ID (number)
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
+외래 키 관계:
+  - singer_in_concert.Singer_ID → singer.Singer_ID
+  - singer_in_concert.concert_ID → concert.concert_ID"""
+    },
+    {
+        "db_id": "pets_1",
+        "question": "가장 나이가 많은 개의 이름은 무엇인가요?",
+        "schema": """데이터베이스 스키마:
+테이블: Student
+컬럼:
+  - StuID (number) (기본 키)
+  - LName (text)
+  - Fname (text)
+  - Age (number)
+  - Sex (text)
+  - Major (number)
+  - Advisor (number)
+  - city_code (text)
+테이블: Has_Pet
+컬럼:
+  - StuID (number)
+  - PetID (number)
+테이블: Pets
+컬럼:
+  - PetID (number) (기본 키)
+  - PetType (text)
+  - pet_age (number)
+  - weight (number)"""
+    },
+    {
+        "db_id": "car_1",
+        "question": "미국산 자동차 중 가장 빠른 자동차는 무엇인가요?",
+        "schema": """데이터베이스 스키마:
+테이블: continents
+컬럼:
+  - ContId (number) (기본 키)
+  - Continent (text)
+테이블: countries
+컬럼:
+  - CountryId (number) (기본 키)
+  - CountryName (text)
+  - Continent (number)
+테이블: car_makers
+컬럼:
+  - Id (number) (기본 키)
+  - Maker (text)
+  - FullName (text)
+  - Country (number)
+테이블: model_list
+컬럼:
+  - ModelId (number) (기본 키)
+  - Maker (number)
+  - Model (text)
+테이블: car_names
+컬럼:
+  - MakeId (number) (기본 키)
+  - Model (text)
+  - Make (text)
+테이블: cars_data
+컬럼:
+  - Id (number) (기본 키)
+  - MPG (text)
+  - Cylinders (number)
+  - Edispl (text)
+  - Horsepower (text)
+  - Weight (number)
+  - Accelerate (number)
+  - Year (number)"""
+    },
+    {
+        "db_id": "tvshow",
+        "question": "가장 높은 평점을 받은 TV 쇼는 무엇인가요?",
+        "schema": """데이터베이스 스키마:
+테이블: TV_Channel
+컬럼:
+  - id (number) (기본 키)
+  - series_name (text)
+  - Country (text)
+  - Language (text)
+  - Content (text)
+  - Pixel_aspect_ratio_PAR (text)
+  - Hight_definition_TV (text)
+  - Pay_per_view_PPV (text)
+  - Package_Option (text)
+테이블: TV_series
+컬럼:
+  - id (number)
+  - Episode (text)
+  - Air_Date (text)
+  - Rating (text)
+  - Share (text)
+  - 18_49_Rating_Share (text)
+  - Viewers_m (text)
+  - Weekly_Rank (number)
+  - Channel (number)
+테이블: Cartoon
+컬럼:
+  - id (number) (기본 키)
+  - Title (text)
+  - Directed_by (text)
+  - Written_by (text)
+  - Original_air_date (text)
+  - Production_code (number)
+  - Channel (number)"""
+    }
+]
+
+def generate_sql(question, db_id, schema_info):
+    """Generate SQL query using the model."""
+    # Create prompt with schema
+    prompt = f"""당신은 자연어 질문을 SQL 쿼리로 변환하는 전문 AI 어시스턴트입니다. 사용자가 데이터베이스에서 정보를 얻기 위해 일상 언어로 질문하면, 당신은 해당 질문을 정확한 SQL 쿼리로 변환해야 합니다.
 
-        response += token
-        yield response
+{schema_info}
 
+질문: {question}
+SQL:"""
+    
+    # 채팅 메시지로 포맷팅
+    messages = [{"role": "user", "content": prompt}]
+    
+    # 채팅 템플릿 적용
+    inputs = tokenizer.apply_chat_template(
+        messages,
+        tokenize=True,
+        add_generation_prompt=True,
+        return_tensors="pt"
+    ).to(model.device)
+    
+    # Generate
+    with torch.no_grad():
+        outputs = model.generate(
+            inputs,
+            max_new_tokens=256,
+            temperature=0.1,
+            top_p=0.95,
+            do_sample=True,
+            use_cache=True
+        )
+    
+    # Decode
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    
+    # Extract SQL after the prompt
+    if prompt in response:
+        sql_part = response.split(prompt)[-1].strip()
+    else:
+        sql_part = response
+    
+    # Clean up the response
+    if sql_part.startswith("assistant"):
+        sql_part = sql_part[len("assistant"):].strip()
+    
+    # Extract SQL query
+    lines = sql_part.split('\n')
+    sql_query = ""
+    for line in lines:
+        line = line.strip()
+        if line.lower().startswith(('select', 'with', '(select')):
+            sql_query = line
+            # Continue collecting lines until we hit a semicolon or empty line
+            for next_line in lines[lines.index(line)+1:]:
+                next_line = next_line.strip()
+                if not next_line or next_line.startswith(('질문', '데이터베이스')):
+                    break
+                sql_query += " " + next_line
+                if next_line.endswith(';'):
+                    break
+            break
+    
+    # Clean up SQL
+    sql_query = sql_query.strip()
+    if sql_query.endswith(';'):
+        sql_query = sql_query[:-1]
+    
+    return sql_query if sql_query else "SQL 생성에 실패했습니다."
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
-)
+def process_question(question, db_id, custom_schema=None):
+    """Process user question and generate SQL query."""
+    if not question or not db_id:
+        return "질문과 데이터베이스 ID를 입력해주세요."
+    
+    # Use custom schema if provided, otherwise find from examples
+    if custom_schema and custom_schema.strip():
+        schema_info = custom_schema
+    else:
+        # Find schema from examples
+        schema_info = None
+        for example in examples:
+            if example["db_id"] == db_id:
+                schema_info = example["schema"]
+                break
+        
+        if not schema_info:
+            return "스키마 정보를 찾을 수 없습니다. 커스텀 스키마를 입력해주세요."
+    
+    # Generate SQL
+    try:
+        sql_query = generate_sql(question, db_id, schema_info)
+        return sql_query
+    except Exception as e:
+        return f"오류 발생: {str(e)}"
 
+# Create Gradio interface
+with gr.Blocks(title="Spider SQL Generator - Korean", theme=gr.themes.Soft()) as demo:
+    gr.Markdown("""
+    # 🕷️ Spider SQL Generator - Korean
+    
+    한국어 질문을 SQL 쿼리로 변환하는 Llama 3.1 8B 모델입니다.
+    
+    ## 📊 성능
+    - **Exact Match**: 42.65%
+    - **Execution Accuracy**: 65.47%
+    - **Training**: Spider 데이터셋 (한국어 번역)
+    """)
+    
+    with gr.Row():
+        with gr.Column():
+            db_id_input = gr.Textbox(
+                label="데이터베이스 ID",
+                placeholder="예: department_management",
+                value="department_management"
+            )
+            
+            question_input = gr.Textbox(
+                label="질문 (한국어)",
+                placeholder="예: 각 부서별 직원 수를 보여주세요.",
+                lines=2
+            )
+            
+            with gr.Accordion("스키마 정보 (선택사항)", open=False):
+                schema_input = gr.Textbox(
+                    label="커스텀 스키마",
+                    placeholder="커스텀 스키마를 입력하세요. 비워두면 예제 스키마를 사용합니다.",
+                    lines=10
+                )
+            
+            submit_btn = gr.Button("SQL 생성", variant="primary", size="lg")
+        
+        with gr.Column():
+            output = gr.Textbox(
+                label="생성된 SQL 쿼리",
+                lines=4,
+                elem_classes=["code"]
+            )
+            
+            gr.Markdown("""
+            ### 💡 사용 팁
+            - 데이터베이스 ID는 예제에서 선택하거나 직접 입력하세요
+            - 질문은 한국어로 자연스럽게 작성하세요
+            - 스키마 정보는 선택사항입니다
+            """)
+    
+    # Examples
+    gr.Markdown("### 📚 예제 (클릭하여 사용)")
+    gr.Examples(
+        examples=[
+            [ex["db_id"], ex["question"], ex["schema"]] for ex in examples
+        ],
+        inputs=[db_id_input, question_input, schema_input],
+        outputs=output,
+        fn=process_question,
+        cache_examples=False
+    )
+    
+    # Submit action
+    submit_btn.click(
+        fn=process_question,
+        inputs=[question_input, db_id_input, schema_input],
+        outputs=output
+    )
+    
+    # Model info
+    gr.Markdown(f"""
+    ---
+    ### 🤖 모델 정보
+    - **Hugging Face**: [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME})
+    - **Base Model**: Llama 3.1 8B
+    - **Fine-tuning**: LoRA with Unsloth
+    - **Dataset**: Spider (Korean translated)
+    """)
 
+# Launch the app
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch(share=True)