FineTurnX/scripts/init_project.py

"""
FineTuneX 项目初始化脚本
"""

import os
import sys


def create_directories():
    """创建必要的目录"""
    directories = [
        "outputs",
        "outputs/checkpoints",
        "data",
        "data/processed",
        "logs",
        "models",
    ]
    
    for directory in directories:
        os.makedirs(directory, exist_ok=True)
        print(f"✓ 创建目录：{directory}")


def create_gitignore():
    """创建 .gitignore 文件"""
    gitignore_content = """# Python
__pycache__/
*.py[cod]
*$py.class
*.so
.Python
build/
develop-eggs/
dist/
downloads/
eggs/
.eggs/
lib/
lib64/
parts/
sdist/
var/
wheels/
*.egg-info/
.installed.cfg
*.egg

# Virtual environments
venv/
env/
ENV/
.venv

# IDE
.vscode/
.idea/
*.swp
*.swo
*~

# Jupyter Notebook
.ipynb_checkpoints

# Project specific
outputs/
*.pth
*.pt
*.bin
*.onnx
logs/
*.log
.DS_Store
Thumbs.db

# Secrets
.env
*.key
*.pem
"""
    
    with open(".gitignore", "w", encoding="utf-8") as f:
        f.write(gitignore_content)
    print("✓ 创建 .gitignore")


def create_env_example():
    """创建 .env.example 文件"""
    env_content = """# HuggingFace
HF_TOKEN=your_huggingface_token_here
HF_ENDPOINT=https://huggingface.co

# Weights & Biases (可选)
WANDB_API_KEY=your_wandb_key_here
WANDB_PROJECT=finetunex

# API Configuration
API_HOST=0.0.0.0
API_PORT=8000
"""
    
    with open(".env.example", "w", encoding="utf-8") as f:
        f.write(env_content)
    print("✓ 创建 .env.example")


def create_readme():
    """创建快速开始 README"""
    readme_content = """# FineTuneX 快速开始

## 1. 安装依赖

```bash
pip install -r requirements.txt
```

## 2. 运行示例

```bash
python examples/qwen3.5_0.8b_finetune.py
```

## 3. 使用自己的数据

准备数据文件 `data.json`:

```json
[
  {
    "instruction": "你的指令",
    "input": "输入（可选）",
    "output": "期望输出"
  }
]
```

修改示例脚本中的数据集路径，然后运行。

## 4. 推理

```bash
python scripts/inference.py --model_path ./outputs/qwen3.5-0.5b-finetuned --interactive
```

## 5. API 服务

```bash
python scripts/start_api.py
```

访问 http://localhost:8000/docs 查看 API 文档。

## 更多信息

查看 [完整文档](docs/usage.md)
"""
    
    with open("QUICKSTART.md", "w", encoding="utf-8") as f:
        f.write(readme_content)
    print("✓ 创建 QUICKSTART.md")


def main():
    print("=" * 60)
    print("FineTuneX 项目初始化")
    print("=" * 60)
    
    create_directories()
    create_gitignore()
    create_env_example()
    create_readme()
    
    print("\n" + "=" * 60)
    print("项目初始化完成！")
    print("=" * 60)
    print("\n下一步:")
    print("1. 查看 QUICKSTART.md 了解快速开始")
    print("2. 查看 docs/usage.md 了解详细文档")
    print("3. 运行：python examples/qwen3.5_0.8b_finetune.py")
    print("=" * 60)


if __name__ == "__main__":
    main()