Simple 2D Navigation task environment for meta reinforcement learning research. Implemented from the paper: Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
Use this code to test the environment.
import gym_2d
env = gym_2d.gym_2d()
observation = env.reset()
print('START:- goal:', env.goal)
for i in range(1000):
env.render()
action = env.sample_action()
observation, reward, done = env.step(action)
if done:
print('DONE:- reached:', observation,',within dist:', env.dist, ',in steps:', i)
break
observation = env.reset()
print('END:- reached:', observation,',within dist:', env.dist, ',in steps:', i)