NousResearch/atropos
1
0
Fork 0
mirror of https://github.com/NousResearch/atropos.git synced 2026-04-19 12:57:58 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
atropos/environments/community/pokemon-showdown
History
crStiv e934094173
Update helpers.py
2025-06-19 22:52:43 +02:00
..
__init__.py Fix final line length violations in Pokemon Showdown environment 2025-05-26 10:15:32 +10:00
helpers.py Update helpers.py 2025-06-19 22:52:43 +02:00
pokemon_env_self_play.py Fix final line length violations in Pokemon Showdown environment 2025-05-26 10:15:32 +10:00
pokemon_environment.py Fix final line length violations in Pokemon Showdown environment 2025-05-26 10:15:32 +10:00
README.md Fix final line length violations in Pokemon Showdown environment 2025-05-26 10:15:32 +10:00
requirements.txt Fix final line length violations in Pokemon Showdown environment 2025-05-26 10:15:32 +10:00

README.md

Pokemon Showdown Environment

This is a Pokemon Showdown environment that allows you to play Pokemon battles with the Pokemon Showdown Battle Simulator.

Overview Video

Video: TBD

Quickstart

  1. Set up the Pokemon Showdown Battle Simulator environment. This is necessary for the pokemon players to be able to connect to the battle simulator.
git clone https://github.com/smogon/pokemon-showdown.git
cd pokemon-showdown
npm install
node pokemon-showdown start --no-security
  1. Run the RL trainer
python environments/game_environments/pokemon-showdown/pokemon_environment.py process

TLDR: GPT player vs Max Damage Player

  • GPT player is a player that uses a GPT model to decide what to do in a battle.    - GPT player received battle history state at the end (binary win/loss) as feedback to the Atropos RL environment.
  • Max Damage Player is a player that always pick the maximum damage move.

Wandb Run

Wandb Run: https://wandb.ai/ajayuppili/atropos-environments_game_environments_pokemon-showdown?nw=nwuserajayuppili