envs.md

envs

ToriLLE comes with OpenAI Gym environment and several pre-defined tasks, which can be used as a (almost) drop-in replacements to e.g. Atari environments.

Register these environments by importing torille.envs.

Limitations:

• seed() function is not implemented (can't change seed of Toribash)
• render() function is not implemented (only one type of state available)

Additional functions/modifications for all environments:

• set_game_draw(draw): Enables/Disables rendering of the game according to boolean parameter.
• settings variable: This is ToribashState object used to set Toribash's settings on each reset.

Solo environments torille.envs.solo_envs.SoloToriEnv

These are tasks where only one character exists (observations/actions only include one player).

Player 2 is set to be immobile and engagement distance is set high to avoid contact between players.

States: 1D vector of player 1 body part positions (gym.spaces.box.Box).

Actions: Joint states for player 1 (gym.spaces.multi_discrete.MultiDiscrete).

Reward is specified by the task (see below).

Settings:

Setting	Value
matchframes	1000
turnframes	5
engagement_distance	1500

Toribash-RunAway-v0

Reward function: Positive reward for head body-part moving away from the center. See torille.envs.solo_envs.reward_run_away.

Toribash-SelfDestruct-v0

Reward function: Positive reward for damaging the player itself (not the opponent). See torille.envs.solo_envs.reward_self_destruct.

Toribash-StaySafe-v0

Reward function: Negative reward for damaging the player itself (not the opponent). See torille.envs.solo_envs.reward_stay_safe.

Uke environments torille.envs.uke_envs.UkeToriEnv

Tasks where only one character is controlled by agent, but observations for both characters are provided.

Player 2 is set to be immobile or random, depending on the task.

States: 1D vector of player 1 and player 2 body part positions (gym.spaces.box.Box)

Actions: Joint states for player 1 (gym.spaces.multi_discrete.MultiDiscrete)

Reward is specified by the task (see below).

Settings:

Setting	Value
matchframes	1000
turnframes	5

Toribash-DestroyUke-v0

Reward function: Positive reward for damaging immobile opponent. See torille.envs.uke_envs.reward_destroy_uke.

Toribash-DestroyUke-v1

Reward function: Positive reward for damaging immobile opponent and negative reward for receiving damage, summed together. See torille.envs.uke_envs.reward_destroy_uke_with_penalty.

Toribash-DestroyUke-v2

Reward function: Positive reward for damaging opponent and negative reward for receiving damage, summed together. Opponent takes random actions each turn. See torille.envs.uke_envs.reward_destroy_uke_with_penalty.

Duo environments torille.envs.duo_envs.DuoToriEnv

Tasks where both characters are controlled by agent and observations are provided for both characters.

States: 1D vector of player 1 and player 2 body part positions (gym.spaces.box.Box)

Actions: Joint states for player 1 and player 2 (gym.spaces.multi_discrete.MultiDiscrete)

Reward is specified by the task (see below).

Settings:

Setting	Value
matchframes	1000
turnframes	5

Toribash-DuoCombat-v0

Reward function: Score from the point of view of player 1: Positive reward if opponent received damage, negative if player 1 received damage (summed together). See torille.envs.duo_envs.reward_player1_pov.

Toribash-Cuddles-v0

Reward function: Positive reward relative to inverse of distance between two players (distance of center-of-masses). Negative reward if either of players takes damage. These are summed together for final reward. See torille.envs.duo_envs.reward_cuddles.

Setting	Value
turnframes	2