TensorTrade is a Python framework for training reinforcement learning agents to trade financial markets.

Build trading bots that LEARN from market data — not just follow fixed rules.

Features:

• Deep RL integration (PPO, DQN, A2C via Ray/RLlib, Stable Baselines)

• Custom reward functions (Sharpe, Sortino, profit factor)

• Multi-asset environments (stocks, crypto, forex)

• Action schemes: simple/managed orders

• Observation pipelines: technical indicators, price data, custom features

• Built-in performance visualization

Stop writing fixed strategies that break when the market shifts. Train agents that adapt.