def main(): logging.info("test xworld navigation goal functions") map_config_files = ['./xworld/map_examples/example7.json'] ego_centrics = [True] for map_config_file in map_config_files: for ego_centric in ego_centrics: args = xworld_args.parser().parse_args() args.ego_centric = ego_centric args.map_config = map_config_file args.show_frame = False env = xworld_navi_goal.XWorldNaviGoal(args) for i in range(2): #2 env.reset() for j in range(200): #20 action = env.agent.random_action() next_state, teacher, done = env.step(action) env.display() if done: logging.info( "test world navigation goal functions done") break logging.info("test world navigation goal functions done")
lr_schedule: schedule for learning rate """ # initialize self.initialize() # model self.evaluate() if __name__ == '__main__': # make env args = xworld_args.parser().parse_args() args.visible_radius_unit_side = config.visible_radius_unit_side args.visible_radius_unit_front = config.visible_radius_unit_front args.ego_centric = config.ego_centric args.map_config = config.map_config_file env = xworld_navi_goal.XWorldNaviGoal(args) env.teacher.israndom_goal = False env.teacher.goal_id = 0 # exploration strategy exp_schedule = LinearExploration(env, config.eps_begin, config.eps_end, config.eps_nsteps) # learning rate schedule lr_schedule = LinearSchedule(config.lr_begin, config.lr_end, config.lr_nsteps) # train model model = DRQN(env, config) shutil.copyfile('./configs/drqn_xworld.py', config.output_path+'config.py') shutil.copy(os.path.realpath(__file__), config.output_path)