#!/usr/bin/env python3
"""
Setup script for MCP Local LLM Server
"""
import os
import sys
import subprocess
import shutil
from pathlib import Path
def check_python_version():
"""Check if Python version is compatible"""
if sys.version_info < (3, 8):
print("❌ Python 3.8 or higher is required")
sys.exit(1)
print(f"✅ Python {sys.version_info.major}.{sys.version_info.minor} detected")
def check_gpu():
"""Check GPU availability"""
try:
import torch
if torch.cuda.is_available():
gpu_count = torch.cuda.device_count()
gpu_name = torch.cuda.get_device_name(0)
print(f"✅ GPU detected: {gpu_name} ({gpu_count} device(s))")
return True
else:
print("⚠️ No GPU detected - will use CPU (slower)")
return False
except ImportError:
print("ℹ️ PyTorch not installed yet - GPU check will happen after installation")
return False
def install_dependencies():
"""Install Python dependencies"""
print("\n📦 Installing dependencies...")
try:
subprocess.check_call([
sys.executable, "-m", "pip", "install", "-r", "requirements.txt"
])
print("✅ Dependencies installed successfully")
except subprocess.CalledProcessError as e:
print(f"❌ Failed to install dependencies: {e}")
sys.exit(1)
def create_env_file():
"""Create .env file from template"""
env_file = Path(".env")
config_file = Path("config.env")
if not env_file.exists() and config_file.exists():
shutil.copy(config_file, env_file)
print("✅ Created .env file from template")
elif env_file.exists():
print("ℹ️ .env file already exists")
else:
print("⚠️ No config template found")
def check_model_cache():
"""Check if model is already cached"""
from transformers import AutoTokenizer
model_name = os.getenv("MODEL_NAME", "Qwen/Qwen3-8B")
try:
# This will download the model if not cached
print(f"\n🔄 Checking model cache for {model_name}...")
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
print(f"✅ Model {model_name} is available")
except Exception as e:
print(f"⚠️ Model download may be needed: {e}")
def main():
"""Main setup function"""
print("🚀 MCP Local LLM Server Setup")
print("=" * 40)
# Check Python version
check_python_version()
# Check GPU
has_gpu = check_gpu()
# Install dependencies
install_dependencies()
# Create environment file
create_env_file()
# Load environment
try:
from dotenv import load_dotenv
load_dotenv()
except ImportError:
pass
# Check model availability
try:
check_model_cache()
except Exception as e:
print(f"⚠️ Could not check model cache: {e}")
print("\n🎉 Setup completed!")
print("\nNext steps:")
print("1. Configure your .env file if needed")
print("2. Test the server: python test_server.py")
print("3. Run the server: python mcp_server.py")
print("4. Integrate with Claude Desktop using claude_desktop_config.json")
if not has_gpu:
print("\n💡 Note: Consider using a smaller model or enabling GPU acceleration for better performance")
if __name__ == "__main__":
main()
