gogo2/crypto/gogo2

Crypto Trading Bot with MEXC API Integration

This is an AI-powered cryptocurrency trading bot that can run in both simulation (demo) mode and live trading mode using the MEXC exchange API.

Features

• Deep Reinforcement Learning agent for trading decisions
• Technical indicators and price prediction
• Live trading integration with MEXC exchange via mexc-api
• Demo mode for testing without real trades
• Real-time data streaming via websockets
• Performance tracking and visualization

Setup

1. Clone the repository
2. Install dependencies:

   pip install -r requirements.txt
   

3. Create a .env file in the root directory with your MEXC API keys:

   MEXC_API_KEY=your_api_key_here
   MEXC_SECRET_KEY=your_secret_key_here
   

Usage

The bot can be run in three different modes:

Training Mode

Train the agent on historical data:

python main.py --mode train --episodes 100

Evaluation Mode

Evaluate the trained agent on historical data:

python main.py --mode evaluate

Live Trading Mode

Run the bot in live trading mode:

python main.py --mode live

To run in demo mode (no real trades):

python main.py --mode live --demo

Live Trading Implementation

The bot uses the mexc-api package to execute trades on the MEXC exchange. The implementation includes:

• Market order execution for opening and closing positions
• Stop loss and take profit orders
• Real-time balance updates
• Trade history tracking

Configuration

You can adjust the following parameters in main.py:

• INITIAL_BALANCE: Starting balance for simulation
• MAX_LEVERAGE: Leverage to use for trading
• STOP_LOSS_PERCENT: Default stop loss percentage
• TAKE_PROFIT_PERCENT: Default take profit percentage

Architecture

• main.py: Main entry point and trading logic
• mexc_trading.py: MEXC API integration for live trading using mexc-api
• models/: Directory for saved model weights

Warning

Trading cryptocurrencies involves significant risk. This bot is provided for educational purposes only. Use at your own risk.

License

MIT