---
language: en
tags:
- reinforcement-learning
- stable-baselines3
- ppo
- trading-bot
- finance
- cryptocurrency
---

# GPy-Trade Generalist PPO Trading Agent

A hierarchical multi-asset generalist reinforcement learning agent for cryptocurrency trading built on top of Stable-Baselines3 (PPO), optimized with pre-trained LSTM trend classifiers, MLP volatility regressors, and an Isolation Forest anomaly detector.

## Model Details
* **Framework**: PyTorch & Gymnasium
* **Algorithm**: PPO (Stable-Baselines3)
* **Assets**: BTCUSDT, ETHUSDT, SOLUSDT, BNBUSDT, XRPUSDT
* **Base Models**:
  * **Classifier**: LSTM (Price direction prediction)
  * **Regressor**: MLP (Future return volatility prediction)
  * **Anomaly Detector**: Isolation Forest (Volatility filter)
* **Observations**: Standardized z-score technical indicators (RSI, MACD, Bollinger Bands, ATR, SMAs) and Cumulative Volume Delta (CVD) to capture order book flow, plus asset-specific `symbol_id`.

## Files included
* `ppo_trading_agent.zip`: Trained PPO model weights
* `scaler.pkl`: StandardScaler fitted on base features
* `anomaly_detector.pkl`: Isolation Forest weights
* `classifier.pt`: PyTorch LSTM direction classifier
* `regressor.pt`: PyTorch MLP volatility regressor