UT-Austin-RPL/poke-env: fork of poke-env that restores original gymnasium envs · GitHub

Name		Name	Last commit message	Last commit date
.github		.github
diagnostic_tools		diagnostic_tools
docs		docs
examples		examples
fixture_data		fixture_data
integration_tests		integration_tests
scripts		scripts
src/poke_env		src/poke_env
unit_tests		unit_tests
.flake8		.flake8
.gitignore		.gitignore
.pre-commit-config.yaml		.pre-commit-config.yaml
.readthedocs.yaml		.readthedocs.yaml
CODE_OF_CONDUCT.md		CODE_OF_CONDUCT.md
CONTRIBUTING.md		CONTRIBUTING.md
LICENSE		LICENSE
MANIFEST.in		MANIFEST.in
README.md		README.md
codecov.yml		codecov.yml
conftest.py		conftest.py
out.json		out.json
pyproject.toml		pyproject.toml
requirements-dev.txt		requirements-dev.txt
requirements.txt		requirements.txt
rl-gif.gif		rl-gif.gif

Repository files navigation

Important

This is a fork of poke-env that installs with metamon. It attempts to extend the lifespan of poke-env as it was during Metamon's development:

Maintains the original gymnasium interface that existed until v0.8.3. OpenAIGymEnv (+ ability to swap in custom Players). Rewards functions that take last_battle and current_battle as input (+ a speed boost). Removes "observation" system that slows fps and is already handled by Metamon.
Preserves minor early-generation battle details as they were when Metamon's original models were trained.
Tries to bring key fixes/improvements since v0.8.3 that are unrelated to gymnasium.

Please see the main repo here for any other use case. I only plan to update this to fix breaking changes to the Showdown sim/request message API. Any improvements to early-generation state tracking/sim protocol are now done in metamon.

Citing poke-env

@misc{poke_env,
    author       = {Haris Sahovic},
    title        = {Poke-env: pokemon AI in python},
    url          = {https://github.com/hsahovic/poke-env}
}