Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Cloud-native Financial Reinforcement Learning
| Date | Stars |
|---|---|
| 2026-07-31 | 506 |
| 2026-08-01 | 506 |
| 2026-08-02 | 506 |
| 2026-08-06 | 506 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# Podracer **News**: We are out of hands, please star it and let us know it is urgent to update this project. Thanks for your feedback. This project can be regarded as [**FinRL**](https://github.com/AI4Finance-LLC/FinRL-Library) **2.0**: intermediate-level framework for full-stack developers and professionals. It is built on [ElegantRL](https://github.com/AI4Finance-LLC/ElegantRL) and [FinRL](https://github.com/AI4Finance-LLC/FinRL-Library) We maintain an **elegant (lightweight, efficient and stable)** FinRL lib, helping researchers and quant traders to develop algorithmic strategies easily. + **Lightweight**: The core codes are less than 800 lines and are based on PyTorch and NumPy. + **Efficient**: Its performance is comparable with [Ray RLlib](https://github.com/ray-project/ray). + **Stable**: It is as stable as [Stable Baseline 3](https://github.com/DLR-RM/stable-baselines3). # Design Principles + **Be Pythonic**: Quant traders, data scientists and machine learning engineers are familiar with the open source Python ecosystem: its programming model, and its tools, e.g., NumPy. + **Put researchers and algorithmic traders first**: Based on PyTorch, we support researchers to mannually control the execution of the codes, empowering them to improve the performance over automatical libraries. + **Lean development of algorithmic strategies**: It is better to have an elegant (may be slightly incomplete) solution than a comprehensive but complex and hard to follow design, e.g., Ray RLlib [link](https://github.com/ray-project/ray). It allows fast code iteration. # DRL Algorithms Currently, most model-free deep reinforcement learning (DRL) algorithms are supported: + **DDPG, TD3, SAC, A2C, PPO, PPO(GAE)** for continuous actions + **DQN, DoubleDQN, D3QN** for discrete actions + **MILP (learn to cut using natural evolutionay strategy)** for portfolio optimization For DRL algorithms, please check out the educational webpage [OpenAI Spinning Up](https://spinningup.openai.com/en/latest/). # File Structure <a href="https://github.com/AI4Finance-LLC/Elegant-FinRL" target="\_blank"> <div align="center"> <img src="https://github.com/Yonv1943/ElegantRL/blob/master/figs/File_structure.png" width="100%"/> </div> <!-- <div align="center"><caption>Slack Invitation Link</caption></div> --> </a> An agent in **agent.py** uses networks in **net.py** and is trained in **run.py** by interacting with an environment in **env.py**. + **net.py** # Neural networks. + Q-Net, + Actor Network, + Critic Network, + **agent.py** # RL algorithms. + AgentBase + AgentDQN + AgentDDPG + AgentTD3 + AgentSAC + AgentPPO + **env.py** # Stock Trading environment + **run.py** # Stock Trading application + Parameter initialization, + Training loop, + Evaluator. + **StockTrading_Demo.ipynb** # One stock trading demo using PPO algorithm # Formulation of the Stock Trading Problem <a href="https://github.com/AI4Finance-LLC/Elegant-FinRL" target="\_blank"> <div align="center"> <img src="figs/1.png" width="50%"/> </div> <!-- <div align="center"><caption>Slack Invitation Link</caption></div> --> </a> Formally, we model stock trading as a Markov Decision Process (MDP), and formulate the trading objective as maximization of expected return: + **State s = [b, p, h]**: a vector that includes the remaining balance b, stock prices p, and stock shares h. p and h are vectors with D dimension, where D denotes the number of stocks. + **Action a**: a vector of actions over D stocks. The allowed actions on each stock include selling, buying, or holding, which result in decreasing, increasing, or no change of the stock shares in h, respectively. + **Reward r(s, a, s’)**: The asset value change of taking action a at state s and arriving at new state s’. + **Policy π(s)**: The trading strategy at state s, which is a probability distribution over actions at state s. + **Q-function Q
Excerpt of 7,138 characters
Read on GitHub
Steven Li
115
Yanglet · United States
44
YonV1943 曾伊言 · SIAT(中科院深圳先进院)
27
Ming Zhu
2
ByFinTech · AI4Finance Foundation Inc. · United States
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:475e2a67f44dffc6, topic:deep-reinforcement-learning, topic:openai-gym, desc:reinforcement learning
matched fp:475e2a67f44dffc6, topic:quantitative-finance, desc:financial