AI Hedge Fund 5zu provides a complete pipeline for quantitative trading: a customizable environment for simulating multiple asset classes, reinforcement learning–based agent modules, backtesting utilities, real-time market data integration, and risk management tools. Users can configure data sources, define reward functions, train agents on historical data, and evaluate performance across key financial metrics. The framework supports modular strategy development and can be extended to live broker APIs for deploying production-level trading bots.

AI Agents for Rock Paper Scissors is an open-source Python project that demonstrates how to build, train, and evaluate different AI strategies—random play, rule-based pattern recognition, and reinforcement learning (Q-learning)—in the classic Rock-Paper-Scissors game. It provides modular agent classes, a configurable game runner, performance logging, and visualization utilities. Users can easily swap agents, adjust learning parameters, and explore AI behavior in competitive scenarios.

The Beer Game Environment provides a discrete-time simulation of a four-stage beer supply chain—retailer, wholesaler, distributor, and manufacturer—exposing an OpenAI Gym interface. Agents receive observations including on-hand inventory, pipeline stock, and incoming orders, then output order quantities. The environment computes per-step costs for inventory holding and backorders, and supports customizable demand distributions and lead times. It integrates seamlessly with popular RL libraries like Stable Baselines3, enabling researchers and educators to benchmark and train algorithms on supply chain optimization tasks.

BotPlayers is a versatile open-source framework designed to streamline the development and deployment of AI-driven game-playing agents. It features a flexible environment abstraction layer that supports screen scraping, web APIs, or custom simulation interfaces, allowing bots to interact with various games. The framework includes built-in reinforcement learning algorithms, genetic algorithms, and rule-based heuristics, along with tools for data logging, model checkpointing, and performance visualization. Its modular plugin system enables developers to customize sensors, actions, and AI policies in Python or Java. BotPlayers also offers YAML-based configuration for rapid prototyping and automated pipelines for training and evaluation. With cross-platform support on Windows, Linux, and macOS, this framework accelerates experimentation and production of intelligent game agents.

CityLearn provides a modular simulation platform for energy management research using reinforcement learning. Users can define multi-zone building clusters, configure HVAC systems, storage units, and renewable sources, then train RL agents against demand response events. The environment exposes state observations like temperatures, load profiles, and energy prices, while actions control setpoints and storage dispatch. A flexible reward API allows custom metrics—such as cost savings or emission reductions—and logging utilities support performance analysis. CityLearn is ideal for benchmarking, curriculum learning, and developing novel control strategies in a reproducible research framework.

CryptoTrader Agents provides a comprehensive toolkit for designing, training, and deploying AI-driven trading strategies in cryptocurrency markets. It includes a modular environment for data ingestion, feature engineering, and custom reward functions. Users can leverage preconfigured reinforcement learning algorithms or integrate their own models. The platform offers simulated backtesting on historical price data, risk management controls, and detailed metric tracking. When ready, agents can connect to live exchange APIs for automated execution. Built on Python, the framework is fully extensible, enabling users to prototype new tactics, run parameter sweeps, and monitor performance in real time.

Fast Reinforcement Learning is a specialized Python framework designed to accelerate the development and execution of reinforcement learning agents. It offers out-of-the-box support for popular algorithms such as PPO, A2C, DDPG and SAC, combined with high-throughput vectorized environment management. Users can easily configure policy networks, customize training loops and leverage GPU acceleration for large-scale experiments. The library’s modular design ensures seamless integration with OpenAI Gym environments, enabling researchers and practitioners to prototype, benchmark and deploy agents across a variety of control, game and simulation tasks.

DeepSeek R1 represents a significant breakthrough in artificial intelligence, delivering top-tier performance in reasoning, mathematics, and coding tasks. Utilizing a sophisticated MoE (Mixture of Experts) architecture with 37B activated parameters and 671B total parameters, DeepSeek R1 implements advanced reinforcement learning techniques to achieve state-of-the-art benchmarks. The model offers robust performance, including 97.3% accuracy on MATH-500 and a 96.3% percentile ranking on Codeforces. Its open-source nature and cost-effective deployment options make it accessible for a wide range of applications.

Dino Reinforcement Learning offers a comprehensive toolkit for training an AI agent to play the Chrome dinosaur game via reinforcement learning. By integrating with a headless Chrome instance through Selenium, it captures real-time game frames and processes them into state representations optimized for deep Q-network inputs. The framework includes modules for replay memory, epsilon-greedy exploration, convolutional neural network models, and training loops with customizable hyperparameters. Users can monitor training progress via console logs and save checkpoints for later evaluation. Post-training, the agent can be deployed to play live games autonomously or benchmarked against different model architectures. The modular design allows easy substitution of RL algorithms, making it a flexible platform for experimentation.

DQN-Deep-Q-Network-Atari-Breakout-TensorFlow provides a complete implementation of the DQN algorithm tailored for the Atari Breakout environment. It uses a convolutional neural network to approximate Q-values, applies experience replay to break correlations between sequential observations, and employs a periodically updated target network to stabilize training. The agent follows an epsilon-greedy policy for exploration and can be trained from scratch on raw pixel input. The repository includes configuration files, training scripts to monitor reward growth over episodes, evaluation scripts to test trained models, and TensorBoard utilities for visualizing training metrics. Users can adjust hyperparameters such as learning rate, replay buffer size, and batch size to experiment with different setups.

Emergent Communication in Agents is an open-source PyTorch framework designed for researchers exploring how multi-agent systems develop their own communication protocols. The library offers flexible implementations of cooperative reinforcement learning tasks, including referential games, combination games, and object identification challenges. Users define speaker and listener agent architectures, specify message channel properties like vocabulary size and sequence length, and select training strategies such as policy gradients or supervised learning. The framework includes end-to-end scripts for running experiments, analyzing communication efficiency, and visualizing emergent languages. Its modular design allows easy extension with new game environments or custom loss functions. Researchers can reproduce published studies, benchmark new algorithms, and probe compositionality and semantics of emergent agent languages.

Gym-Recsys is a toolbox that wraps recommendation tasks into OpenAI Gym environments, allowing reinforcement learning algorithms to interact with simulated user-item matrices step by step. It provides synthetic user behavior generators, supports loading popular datasets, and delivers standard recommendation metrics like Precision@K and NDCG. Users can customize reward functions, user models, and item pools to experiment with different RL-based recommendation strategies in a reproducible manner.

GridWorldEnvs offers a comprehensive suite of grid-world environments to support the design, testing, and benchmarking of reinforcement learning and multi-agent systems. Users can easily configure grid dimensions, agent start positions, goal locations, obstacles, reward structures, and action spaces. The library includes ready-to-use templates such as classic grid navigation, obstacle avoidance, and cooperative tasks, while also allowing custom scenario definitions via JSON or Python classes. Seamless integration with the OpenAI Gym API means that standard RL algorithms can be applied directly. Additionally, GridWorldEnvs supports single-agent and multi-agent experiments, logging, and visualization utilities for tracking agent performance.

gym-fx is an open-source Python library that implements a simulated Forex trading environment using the OpenAI Gym interface. It offers support for multiple currency pairs, integration of historical price feeds, technical indicators, and fully customizable reward functions. By providing a standardized API, gym-fx simplifies the process of benchmarking and developing reinforcement learning algorithms for algorithmic trading. Users can configure market slippage, transaction costs, and observation spaces to closely mimic live trading scenarios, facilitating robust strategy development and evaluation.

gym-llm extends the OpenAI Gym ecosystem to large language models by defining text-based environments where LLM agents interact through prompts and actions. Each environment follows Gym’s step, reset, and render conventions, emitting observations as text and accepting model-generated responses as actions. Developers can craft custom tasks by specifying prompt templates, reward calculations, and termination conditions, enabling sophisticated decision-making and conversational benchmarks. Integration with popular RL libraries, logging tools, and configurable evaluation metrics facilitates end-to-end experimentation. Whether assessing an LLM’s ability to solve puzzles, manage dialogues, or navigate structured tasks, gym-llm provides a standardized, reproducible framework for research and development of advanced language agents.

gym-multigrid provides a suite of customizable gridworld environments designed for multi-room navigation and exploration tasks in reinforcement learning. Each environment consists of interconnected rooms populated with objects, keys, doors, and obstacles. Users can adjust grid size, room configurations, and object placements programmatically. The library supports both full and partial observation modes, offering RGB or matrix state representations. Actions include movement, object interaction, and door manipulation. By integrating it as a Gym environment, researchers can leverage any Gym-compatible agent, seamlessly training and evaluating algorithms on tasks like key-door puzzles, object retrieval, and hierarchical planning. gym-multigrid’s modular design and minimal dependencies make it ideal for benchmarking new AI strategies.

HFO_DQN combines Python and TensorFlow to deliver a complete pipeline for training soccer agents using Deep Q-Networks. Users can clone the repository, install dependencies including the HFO simulator and Python libraries, and configure training parameters in YAML files. The framework implements experience replay, target network updates, epsilon-greedy exploration, and reward shaping tailored for the half field offense domain. It features scripts for agent training, performance logging, evaluation matches, and plotting results. Modular code structure allows integration of custom neural network architectures, alternative RL algorithms, and multi-agent coordination strategies. Outputs include trained models, performance metrics, and behavior visualizations, facilitating research in reinforcement learning and multi-agent systems.

jason-RL adds a reinforcement learning layer to the Jason multi-agent framework, allowing AgentSpeak BDI agents to learn action-selection policies via reward feedback. It implements Q-learning and SARSA algorithms, supports configuration of learning parameters (learning rate, discount factor, exploration strategy), and logs training metrics. By defining reward functions in agent plans and running simulations, developers can observe agents improve decision making over time, adapting to changing environments without manual policy coding.

MARFT is a Python-based LLMs, enabling reproducible experiments and rapid prototyping of collaborative AI systems.

MineLand provides a flexible 3D sandbox environment inspired by Minecraft for training reinforcement learning agents. It features Gym-compatible APIs for seamless integration with existing RL libraries such as Stable Baselines, RLlib, and custom implementations. Users gain access to a library of tasks, including resource collection, navigation, and construction challenges, each with configurable difficulty and reward structures. Real-time rendering, multi-agent scenarios, and headless modes allow for scalable training and benchmarking. Developers can design new maps, define custom reward functions, and plugin additional sensors or controls. MineLand’s open-source codebase fosters reproducible research, collaborative development, and rapid prototyping of AI agents in complex virtual worlds.