crowdAI / marLo

Multi Agent Reinforcement Learning using MalmÖ
MIT License
245 stars 46 forks source link

Default Reward, Observation, Video Policy ? #18

Open spMohanty opened 6 years ago

spMohanty commented 6 years ago

@katja-hofmann @AndKram : Suggesstions on the default policies from : https://microsoft.github.io/malmo/0.17.0/Documentation/classmalmo_1_1_agent_host.html

VideoPolicy { LATEST_FRAME_ONLY,  KEEP_ALL_FRAMES  }
RewardsPolicy { LATEST_REWARD_ONLY,  SUM_REWARDS,  KEEP_ALL_REWARDS  }
ObservationsPolicy { LATEST_OBSERVATION_ONLY,  KEEP_ALL_OBSERVATIONS  }