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.
Texas holdem OpenAi gym poker environment with reinforcement learning based on keras-rl. Includes virtual rendering and montecarlo for equity calculation.
| Date | Stars |
|---|---|
| 2026-07-31 | 721 |
| 2026-08-01 | 721 |
| 2026-08-02 | 721 |
| 2026-08-06 | 721 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
Neuron Poker: OpenAi gym environment for texas holdem poker =========================================================== This is an environment for training neural networks to play texas holdem. Please try to model your own players and create a pull request so we can collaborate and create the best possible player. Usage: ------ Run: - Install Python 3.11, I would also recommend to install PyCharm. - Install uv with ``curl -LsSf https://astral.sh/uv/install.sh | sh`` - Create a virtual environment and install dependencies with ``uv sync`` - Run 6 random players playing against each other: ``uv run poker-random-render`` or - To manually control the players: ``uv run poker-keypress-render`` - Example of genetic algorithm with self improvement: ``uv run poker-equity-improvement`` - In order to use the C++ version of the equity calculator, you will also need to install Visual Studio 2019 (or GCC over Cygwin may work as well). To use it, use the -c option when running main.py. - For more advanced users: ``uv run poker-dqn-train-cpp`` will start training the deep Q agent with C++ Monte Carlo for faster calculation - Run all tests: ``uv run pytest`` (use -n to run tests in parallel) .. figure:: doc/table.gif :alt: Analysis of the run ~~~~~~~~~~~~~~~~~~~ At the end of an episode, the performance of the players can be observed via the summary plot. |image0| Packages and modules: ~~~~~~~~~~~~~~~~~~~~~ main.py: entry point and command line interpreter. Runs agents with the gym. The docstring at the top of the file describes the command line options. They are interpreted by docopt. gym\_env ~~~~~~~~ - ``env.py``: Texas Hold’em unlimited openai gym environment & ``rendering.py``: rendering graphics while playing agents ~~~~~~ Please add your model based agents here. - ``agent_random.py``: an agent making random decisions - ``agent_keypress.py``: an agent taking decision via keypress - ``agent_consider_equity.py``: an agent considering equity information - ``agent_keras_rl_dqn.py``: Deep Q learning agent, using keras-rl for deep reinforcement learning - ``agent_custom_q1.py``: Custom implementation of deep q learning Note that the observation property is a dictionary that contains all the information about the players and table that can be used to make a decision. Custom implementation of q learning ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Custom impelemtation of reinforcement learning. This package is now in a separate repo: www.github.com/dickreuter/tf_rl tools ~~~~~ - ``hand_evaluator.py``: evaluate the best hand of multiple players - ``helper.py``: helper functions - ``montecarlo_numpy2.py``: fast numpy based montecarlo simulation to calculate equity. Not yet working correctly. Some tests are failing. Feel free to fix them. - ``montecarlo_python.py``: relatively slow python based montecarlo for equity calculation. Supports preflight ranges for other players. - ``montecarlo_cpp``: c++ implementation of equity calculator. Around 500x faster than python version tests ^^^^^ - ``test_gym_env.py``: tests for the end. - ``test_montecarlo.py``: tests for the hands evaluator and python based equity calculator. - ``test_montecarlo_numpy.py``: tests for the numpy montecarlo Roadmap ------- Agents ~~~~~~ - [x] Agent based on user interaction (keypress) - [x] Random agent - [x] Equity based strategy (i.e. call and bet above threshold) - [x] Equity based strategy with genetic algorithm, adjusting the treshold based on winning agent. - [x] C++ implementation of equity calculator to significantly speed up runs - [x] Agent based on reinforcement learning with experience replay (Deep Q learning, based on keras-rl) - [/] Custom agents (see above section for more details) Reinforcement learning: Deep Q agent ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ``neuron_poker.agents.agent_dqn`` implements a deep q agent with help of keras-rl. A number of parameters can be se: - nb_max_start_steps = 20 # maximum of rando
Excerpt of 10,540 characters
Read on GitHub
69
7
5
5
1
1
1
1
1
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:9ca1e32f321311ba, topic:reinforcement-learning, topic:openai-gym, desc:reinforcement learning