Metadata-Version: 2.1
Name: aleph0
Version: 0.0.10000069
Summary: alphazero implemented for games with state/action spaces a finite subset of a countable space
Author-email: Pranav <pranavraj575@gmail.com>
Project-URL: Homepage, https://github.com/pranavraj575/aleph0
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: torch
Requires-Dist: numpy
Provides-Extra: jenga
Requires-Dist: scipy; extra == "jenga"

# AlephZero: Extending AlphaZero to Infinite Boards

General implementation of $\aleph_0$,
an extension of AlphaZero which uses Transformer network architecture to solve a class of games 
($\aleph_0$-extensive form games) whose observation space is sequential, 
and whose action space arises from choosing indices in this observation space.

## Installation
* Either from pip: ```pip3 install aleph0```
* Or clone from repo:
  ```bash
  git clone https://github.com/pranavraj575/aleph0
  pip3 install -e aleph0
  ```
To play Jenga, must install additional packages ```pip3 install aleph0[jenga]```
