0.2.1

This release introduces two functions to clone the state of the environment:

• get_state: gets the state of the environment (board, holder, ...)
• set_state: sets the state of the environment

These functions are the preferred way to clone a Tetris environment. Compared to deepcopy, this method is a lot faster.

0.2.1-test

This release introduces two functions to clone the state of the environment:

• get_state: gets the state of the environment (board, holder, ...)
• set_state: sets the state of the environment

These functions are the preferred way to clone a Tetris environment. Compared to deepcopy, this method is a lot faster.

0.2.0

This is the first PyPi release of Tetris Gymnasium.

Features

• Supports classic action-space
• Supports grouped-actions action space as a wrapper
• Supports customization for tetrominoes, board dimensions, gravity and rewards via parameters
• Includes wrappers for rendering
• Includes example scripts for training agents (DQN) and logging in W&B
• Includes a Dockerfile
• Offers a detailed documentation on GitHub pages
• Accompanied a short-paper describing the motivation and implementation details

Note: This version may have issues which will be resolved in future versions. Milestones and issues are tracked in the GitHub project.

PyPi Test Prerelease

0.2.0-test

Remove dynamic version and set version to 0.2.0

PyPi Publish Test

0.1.0

Change repo url