Gym Environment for Guessing Multiple Numbers. Inspired by GoodAI and the Badger architecture. For more information on how the guessing game is used in the Badger architecture, see the original badger paper.
clone this repository and install the package:
$git clone https://github.com/jacobbriones1/GuessNumbers
$cd GuessNumbers
$pip install -e .
import gym
import guessing_game
env = gym.make('NumberGuessing-v0')
env.reset(size=5)
The goal is to guess a vector of a given size. A vector is initialized randomly, and an agent guesses using step([...])
. The returned observation is the Mean Squared Error between the actual vector and the guessed vector. A reward of 0 is given for guesses that are not within an MSE of 0.001, and a reward of 1 otherwise.