The bug when convert the demo data?

[yingfei1016]

Hi，

When I convert the demo data to get the RGB video by https://github.com/haosulab/ManiSkill2-Learn/blob/main/scripts/example_demo_conversion/general_rigid_body_single_object_envs.sh . Some tasks will get error. I use the main branch of the ManiSkill2-Learn. For example: 1、TypeError: Wrapper.reset() got an unexpected keyword argument 'episode_idx' command: ` ENV="PandaAvoidObstacles-v0" python -m mani_skill2.trajectory.replay_trajectory --num-procs 4 \ --traj-path demos/v0/rigid_body/$ENV/trajectory.h5 \ --save-traj \ --target-control-mode pd_ee_delta_pose \ --obs-mode none

python tools/convert_state.py \ --env-name=$ENV \ --num-procs=1 \ --traj-name=../ManiSkill2/demos/v0/rigid_body/$ENV/trajectory.none.pd_ee_delta_pose.h5 \ --json-name=../ManiSkill2/demos/v0/rigid_body/$ENV/trajectory.none.pd_ee_delta_pose.json \ --output-name=../ManiSkill2/demos/v0/rigid_body/$ENV/trajectory.none.pd_ee_delta_pose_rgbd.h5 \ --control-mode=pd_ee_delta_pose \ --max-num-traj=-1 \ --obs-mode=rgbd \ --reward-mode=dense ` Error:

Reset kwargs for the current trajectory: {'seed': 0, 'episode_idx': 0} Traceback (most recent call last): File "/home/megvii/omniverse/ov/pkg/ManiSkill2-Learn/tools/convert_state.py", line 254, in <module> main() File "/home/megvii/omniverse/ov/pkg/ManiSkill2-Learn/tools/convert_state.py", line 231, in main convert_state_representation(keys, args, 0, os.getpid(), *extra_args) File "/home/megvii/omniverse/ov/pkg/ManiSkill2-Learn/tools/convert_state.py", line 84, in convert_state_representation env.reset(**reset_kwargs[cur_episode_num]) File "/home/megvii/omniverse/ov/pkg/ManiSkill2-Learn/maniskill2_learn/env/wrappers.py", line 100, in reset obs, _ = self.env.reset(*args, **kwargs) # ignore reset info in gymnasium File "/home/megvii/.local/lib/python3.10/site-packages/gymnasium/wrappers/time_limit.py", line 75, in reset return self.env.reset(**kwargs) File "/home/megvii/omniverse/ov/pkg/ManiSkill2-Learn/maniskill2_learn/env/wrappers.py", line 223, in reset obs, reset_info = self.env.reset(**kwargs) File "/home/megvii/omniverse/ov/pkg/ManiSkill2-Learn/maniskill2_learn/env/wrappers.py", line 692, in reset obs, reset_info = super().reset(**kwargs) TypeError: Wrapper.reset() got an unexpected keyword argument 'episode_idx'

2、TypeError: Wrapper.reset() got an unexpected keyword argument 'model_id' command: ` ENV="PickSingleEGAD-v0" python -m mani_skill2.trajectory.replay_trajectory --num-procs 4 \ --traj-path demos/v0/rigid_body/$ENV/trajectory.h5 \ --save-traj \ --target-control-mode pd_ee_delta_pose \ --obs-mode none

python tools/convert_state.py \ --env-name=$ENV \ --num-procs=1 \ --traj-name=../ManiSkill2/demos/v0/rigid_body/$ENV/trajectory.none.pd_ee_delta_pose.h5 \ --json-name=../ManiSkill2/demos/v0/rigid_body/$ENV/trajectory.none.pd_ee_delta_pose.json \ --output-name=../ManiSkill2/demos/v0/rigid_body/$ENV/trajectory.none.pd_ee_delta_pose_rgbd.h5 \ --control-mode=pd_ee_delta_pose \ --max-num-traj=-1 \ --obs-mode=rgbd \ --reward-mode=dense ` Error:

` Reset kwargs for the current trajectory: {'seed': 0, 'model_id': 'A10_0'} Traceback (most recent call last): File "/home/megvii/omniverse/ov/pkg/ManiSkill2-Learn/tools/convert_state.py", line 254, in main() File "/home/megvii/omniverse/ov/pkg/ManiSkill2-Learn/tools/convert_state.py", line 231, in main convert_state_representation(keys, args, 0, os.getpid(), *extra_args) File "/home/megvii/omniverse/ov/pkg/ManiSkill2-Learn/tools/convert_state.py", line 84, in convert_state_representation env.reset(*reset_kwargs[cur_episode_num]) File "/home/megvii/omniverse/ov/pkg/ManiSkill2-Learn/maniskill2learn/env/wrappers.py", line 100, in reset obs, = self.env.reset(args, kwargs) # ignore reset info in gymnasium File "/home/megvii/.local/lib/python3.10/site-packages/gymnasium/wrappers/time_limit.py", line 75, in reset return self.env.reset(kwargs) File "/home/megvii/omniverse/ov/pkg/ManiSkill2-Learn/maniskill2_learn/env/wrappers.py", line 223, in reset obs, reset_info = self.env.reset(kwargs) File "/home/megvii/omniverse/ov/pkg/ManiSkill2-Learn/maniskill2_learn/env/wrappers.py", line 692, in reset obs, reset_info = super().reset(kwargs) TypeError: Wrapper.reset() got an unexpected keyword argument 'model_id'

`

[xuanlinli17]

PickSingleEGAD-v0 is not a single object task. Please use https://github.com/haosulab/ManiSkill2-Learn/blob/main/scripts/example_demo_conversion/pick_place_multi_object_envs.sh

[xuanlinli17]

Also please note that different scripts have comments indicating the envs they apply.

[yingfei1016]

What scripts should the PandaAvoidObstacles-v0 use?

[xuanlinli17]

https://github.com/haosulab/ManiSkill2-Learn/blob/main/scripts/example_demo_conversion/general_rigid_body_single_object_envs.sh

[yingfei1016]

Can you look at my error again? I just used https://github.com/haosulab/maniskill2-learn/blob/main/scripts/example demo conversion/general rigid body single object _ envs.sh, but I got an error when env reset.

Thank you!

[yingfei1016]

File "/home/megvii/omniverse/ov/pkg/ManiSkill2-Learn/maniskill2_learn/env/wrappers.py", line 692, in reset obs, reset_info = super().reset(**kwargs) TypeError: Wrapper.reset() got an unexpected keyword argument 'episode_idx'

[yingfei1016]

If the 'episode_idx' and 'model_id' are not necessary, I will remove them directly.

[xuanlinli17]

Which commit of ManiSkill2 are you using?

episode_idx and model_id are needed.

[yingfei1016]

I install the ManiSkill2 by pip install, the version is 0.5.3 maniSkill2_learn is 1.8.0b0

[yingfei1016]

I try to use the ManiSkill2==0.5.0, but still has this problem.

[yingfei1016]

@xuanlinli17 If you have any solution to solve this problem?