📋 Prerequisites & Requirements
⚠️ CRITICAL: This tutorial sets up PAPER TRADING ONLY. Never use real money initially.
🛠️ Hardware Requirements
- Raspberry Pi 5 (Recommended): 4GB RAM minimum, 8GB preferred
- MicroSD Card: 32GB+ Class 10 or better
- Power Supply: Official Raspberry Pi power adapter
- Internet Connection: Ethernet or WiFi
- Optional: External SSD for better performance
💻 Alternative Hardware
- Desktop/Laptop: Linux, macOS, or Windows with Python 3.11+
- VPS/Cloud: AWS, Digital Ocean, or similar (always-on)
- Docker: Any system supporting Docker containers
🧠 Knowledge Prerequisites
- Basic Command Line: Navigate directories, edit files
- Text Editing: Use nano, vim, or similar editors
- Understanding of Risk: This is experimental software
- Trading Basics: Know what BUY/SELL/HOLD means
🔑 API Requirements
- Exchange API Keys: Binance, Coinbase Pro, or Kraken
- AI Model APIs: OpenAI, Anthropic, Google (optional)
- Local LLM Option: Ollama for privacy-focused setup
🚀 Installation Guide
Step 1: Prepare Raspberry Pi
# Download Raspberry Pi Imager
# Flash Raspberry Pi OS Lite (64-bit) to microSD
# Enable SSH during imaging process
# Set username: pi
# Set password: [your-secure-password]
# Configure WiFi if needed
Step 2: First Boot Setup
# SSH into your Raspberry Pi
ssh pi@[your-pi-ip-address]
# Update system
sudo apt update && sudo apt upgrade -y
# Install essential packages
sudo apt install -y python3.11 python3.11-venv python3-pip git curl
# Install Docker (optional but recommended)
curl -fsSL https://get.docker.com -o get-docker.sh
sudo sh get-docker.sh
sudo usermod -aG docker pi
Step 3: Clone BerryBotPi Repository
# Create projects directory
mkdir ~/projects && cd ~/projects
# Clone the repository
git clone https://github.com/berrybot-ai/berrybot-mk3.git
cd berrybot-mk3
# Check repository structure
ls -la
# Should see: src/, config/, docs/, requirements.txt, README.md
Step 4: Python Environment Setup
# Create virtual environment
python3.11 -m venv venv
# Activate virtual environment
source venv/bin/activate
# Upgrade pip
pip install --upgrade pip
# Install BerryBotPi dependencies
pip install -r requirements.txt
# Verify installation
python -c "import berrybot; print('BerryBotPi installed successfully')"
Step 5: Install Local LLMs (Optional)
# Install Ollama for local AI models
curl -fsSL https://ollama.ai/install.sh | sh
# Download recommended models (this takes time)
ollama pull llama3.1:8b
ollama pull qwen2.5:7b
ollama pull gemma2:9b
# Verify Ollama installation
ollama list
⚙️ Configuration Setup
✅ SAFETY FIRST: All configuration starts in paper trading mode.
Step 1: Basic Configuration
# Copy example configuration
cp config/config.example.json config/config.json
# Edit configuration file
nano config/config.json
# CRITICAL: Ensure these settings
{
"trading_mode": "paper", # NEVER change this initially
"paper_balance": 10000, # Start with $10,000 fake money
"max_position_size": 0.05, # Max 5% per trade
"stop_loss_percentage": 0.02, # 2% stop loss
"take_profit_percentage": 0.04, # 4% take profit
"safety_checks": true,
"confirm_all_trades": true
}
Step 2: Exchange Configuration
# Edit exchange settings
nano config/exchanges.json
{
"binance": {
"enabled": true,
"api_key": "your-api-key-here",
"api_secret": "your-api-secret-here",
"testnet": true, # Use testnet for safety
"paper_trading": true # Double safety check
}
}
# IMPORTANT: Start with read-only API keys
# Add trading permissions only after extensive paper testing
Step 3: LLM Configuration
# Configure AI models
nano config/llms.json
{
"enabled_llms": [
"ollama_llama3.1", # Local model (free)
"ollama_qwen2.5", # Local model (free)
"openai_gpt4o_mini" # Cloud model (paid)
],
"ollama": {
"base_url": "http://localhost:11434",
"timeout": 120
},
"openai": {
"api_key": "your-openai-api-key",
"model": "gpt-4o-mini", # Cheaper option
"max_tokens": 500
},
"consensus": {
"minimum_votes": 2, # At least 2 models must agree
"confidence_threshold": 0.6 # 60% minimum confidence
}
}
Step 4: Risk Management Setup
# Configure risk parameters
nano config/risk.json
{
"max_daily_trades": 5, # Limit trading frequency
"max_daily_loss": 0.02, # Stop if lose 2% in one day
"max_position_size": 0.05, # Never more than 5% per trade
"blackout_periods": [ # Don't trade during major events
"2024-12-25", # Christmas
"2024-01-01" # New Year
],
"emergency_stop": {
"enabled": true,
"max_consecutive_losses": 3, # Stop after 3 losses in a row
"volatility_threshold": 0.1 # Stop if market too volatile
}
}
Step 5: Logging Configuration
# Set up comprehensive logging
nano config/logging.json
{
"level": "INFO",
"log_to_file": true,
"log_directory": "./logs",
"max_file_size": "10MB",
"backup_count": 30,
"loggers": {
"trading_decisions": "logs/decisions.log",
"ai_responses": "logs/ai_responses.log",
"api_calls": "logs/api_calls.log",
"errors": "logs/errors.log",
"performance": "logs/performance.log"
}
}
# Create logs directory
mkdir -p logs
🧪 Testing & Validation
Step 1: System Validation
# Activate virtual environment
source venv/bin/activate
# Run system checks
python src/berrybot/main.py --validate-config
python src/berrybot/main.py --test-apis
python src/berrybot/main.py --test-llms
# All tests should pass before proceeding
Step 2: Paper Trading Test
# Start paper trading mode
python src/berrybot/main.py --paper-trading
# In another terminal, monitor logs
tail -f logs/decisions.log
# Should see output like:
# [2024-01-08 10:00:01] PAPER MODE: Analyzing BTC/USDT
# [2024-01-08 10:02:15] CONSENSUS: 3/3 models vote HOLD (confidence: 72%)
# [2024-01-08 10:02:16] PAPER TRADE: No action taken
Step 3: Performance Monitoring
# Check paper trading results
python scripts/analyze_performance.py --paper-only
# View trading dashboard (if web interface enabled)
python src/web_ui/app.py
# Visit: http://your-pi-ip:8501
# Daily status check
python scripts/daily_report.py
Step 4: Safety Verification
# Verify paper trading mode is active
python scripts/verify_safety.py
# Output should be:
# ✅ PAPER TRADING MODE: ACTIVE
# ✅ API KEYS: READ-ONLY
# ✅ REAL TRADING: DISABLED
# ✅ SAFETY CHECKS: ENABLED
# If any check fails, DO NOT PROCEED
Step 5: Set Up Monitoring
# Install system service (optional)
sudo cp scripts/berrybot.service /etc/systemd/system/
sudo systemctl daemon-reload
sudo systemctl enable berrybot.service
# Start service
sudo systemctl start berrybot.service
# Check status
sudo systemctl status berrybot.service
# View logs
journalctl -u berrybot.service -f
🎯 Next Steps & Best Practices
📚 LEARNING PHASE: Paper trade for minimum 3 months before considering real money.
Immediate Actions (Week 1)
- Monitor Daily: Check logs and decisions every day
- Understand Votes: Read AI reasoning for each decision
- Track Performance: Record wins/losses in paper trading
- Join Community: Discord, Reddit for questions and support
Short-term Goals (Month 1-3)
- Market Analysis: Learn how AI reacts to different conditions
- Fine-tuning: Adjust parameters based on results
- Emergency Drills: Practice stopping the system quickly
- Backup Systems: Set up monitoring and alerts
Long-term Preparation (Month 3+)
- Performance Review: Analyze 3+ months of data
- Risk Assessment: Understand failure modes
- Decision Making: Determine if system is reliable enough
- Micro Testing: If results are positive, consider tiny real amounts
Red Flags - Stop Immediately If:
- Paper trading loses more than 10% in any month
- System has frequent technical failures
- AI models consistently make poor decisions
- You don't understand why decisions are being made
- Market conditions change significantly
Resources & Support
# Final safety check command
python scripts/final_safety_check.py
# Should output:
# 🔒 BerryBotPi Safety Status Report
# ✅ Paper trading mode: ACTIVE
# ✅ Real trading: DISABLED
# ✅ API permissions: READ-ONLY
# ✅ Emergency stop: CONFIGURED
# ✅ Monitoring: ENABLED
#
# 🎓 READY FOR EDUCATIONAL USE
# ⚠️ NEVER USE REAL MONEY WITHOUT MONTHS OF TESTING