Stuck in eval_seq2seq.py

[HANQAQ]

Hello, when i run eval_seq2seq.py, i found it will stuck in

def setup_scene(cls, env, traj_data, r_idx, args, reward_type='dense'):
        '''
        intialize the scene and agent from the task info
        '''
        # scene setup
        scene_num = traj_data['scene']['scene_num']
        object_poses = traj_data['scene']['object_poses']
        dirty_and_empty = traj_data['scene']['dirty_and_empty']
        object_toggles = traj_data['scene']['object_toggles'] 

        scene_name = 'FloorPlan%d' % scene_num
        env.reset(scene_name) 
        env.restore_scene(object_poses, object_toggles, dirty_and_empty) 

        # initialize to start position
        env.step(dict(traj_data['scene']['init_action']))

        # print goal instr
        print("Task: %s" % (traj_data['turk_annotations']['anns'][r_idx]['task_desc']))

        # setup task for reward
        env.set_task(traj_data, args, reward_type=reward_type)

especially when i run any about env, for example, env.reset(scene_name), env.restore_scene(object_poses, object_toggles, dirty_and_empty)

How can i do to solve this problem :）

[Diginyx]

@HANQAQ did you ever figure this out?

[HANQAQ]

@HANQAQ did you ever figure this out?

Not yet. I try to many time, but it will stuck. :(