initial commit

.gitattributes

@@ -33,3 +33,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+*.msh filter=lfs diff=lfs merge=lfs -text
+*.jpg filter=lfs diff=lfs merge=lfs -text
+*.png filter=lfs diff=lfs merge=lfs -text

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 Zichen Jeff Cui
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README.md

@@ -1,3 +1,140 @@
----
-license: mit
----
+# **DynaMo**: In-Domain Dynamics Pretraining for Visuo-Motor Control
+[[Paper]](https://arxiv.org/abs/2409.12192) [[Project Website]](https://dynamo-ssl.github.io/) [[Data]](https://osf.io/kxehw/)
+
+[Zichen Jeff Cui](https://jeffcui.com/), [Hengkai Pan](https://www.ri.cmu.edu/ri-people/hengkai-pan/), [Aadhithya Iyer](https://aadhithya14.github.io/), [Siddhant Haldar](https://siddhanthaldar.github.io/) and [Lerrel Pinto](https://www.lerrelpinto.com/), New York University
+
+This repo contains code for DynaMo visual pretraining, and for reproducing sim environment experiments. Datasets will be uploaded soon.
+
+
+## Getting started
+The following assumes our current working directory is the root directory of this project repo; tested on Ubuntu 22.04 LTS (amd64).
+### Setting up the project environments
+- Install the project environment:
+  ```
+  conda env create --file=conda_env.yml
+  ```
+- Activate the environment:
+  ```
+  conda activate dynamo-repro
+  ```
+- To enable logging, log in with a `wandb` account:
+  ```
+  wandb login
+  ```
+  Alternatively, to disable logging altogether, set the environment variable `WANDB_MODE`:
+  ```
+  export WANDB_MODE=disabled
+  ```
+
+### Getting the training datasets
+[Get the dataset here](https://osf.io/kxehw/).
+
+(Updated Sep 29: sim kitchen dataset now supports lazy loading: set `prefetch=False` in the sim kitchen configs. If you encounter errors, try downloading the latest dataset zips from the link above.)
+- Download all files in the `datasets` directory, combine all partitions, and unzip:
+  ```
+  zip -s- dynamo_repro_datasets.zip -O combined.zip
+  unzip combined.zip
+  ```
+- In `./configs/env_vars/env_vars.yaml`, set `dataset_root` to the unzipped parent directory containing all datasets.
+- In `./eval_configs/env_vars/env_vars.yaml`, set `dataset_root` to the unzipped parent directory containing all datasets.
+- In `./eval_configs/env_vars/env_vars.yaml`, set `save_path` to where you want to save the rollout results (e.g. root directory of this repo).
+- Environments:
+  - `sim_kitchen`: Franka kitchen environment
+  - `block_push_multiview`: Block push environment
+  - `libero_goal`: LIBERO Goal environment
+  - `pusht`: Push-T environment
+
+## Reproducing experiments
+The following assumes our current working directory is the root directory of this project repo.
+
+To reproduce the experiment results, the overall steps are:
+1. Activate the conda environment with
+   ```
+   conda activate dynamo-repro
+   ```
+
+2. Train the visual encoder with `python3 train.py --config-name=train_*`. A model snapshot will be saved to `./exp_local/...`;
+3. In `eval_configs/encoder`, in the corresponding environment config, set the encoder file path `f` to the saved snapshot;
+4. Eval with `python3 online_eval.py --config-name=train_*`.
+
+See below for detailed steps for each environment.
+
+
+### Franka Kitchen
+- Train the encoder:
+  ```
+  python3 train.py --config-name=train_sim_kitchen
+  ```
+  Snapshots will be saved to a new timestamped directory `./exp_local/{date}/{time}_train_sim_kitchen_dynamo`.
+
+  The encoder snapshot will be at `./exp_local/{date}/{time}_train_sim_kitchen_dynamo/encoder.pt`.
+- In `eval_configs/encoder/kitchen_dynamo.yaml`, set `SNAPSHOT_PATH` to the absolute path of the encoder snapshot above.
+- Evaluation:
+  ```
+  MUJOCO_GL=egl python3 online_eval.py --config-name=train_sim_kitchen
+  ```
+
+### Block Pushing
+- Train the encoder:
+  ```
+  python3 train.py --config-name=train_blockpush
+  ```
+  Snapshots will be saved to a new timestamped directory `./exp_local/{date}/{time}_train_blockpush_dynamo`.
+
+  The encoder snapshot will be at `./exp_local/{date}/{time}_train_blockpush_dynamo/encoder.pt`.
+- In `eval_configs/encoder/blockpush_dynamo.yaml`, set `SNAPSHOT_PATH` to the absolute path of the encoder snapshot above.
+- Evaluation:
+  ```
+  ASSET_PATH=$(pwd) python3 online_eval.py --config-name=train_blockpush
+  ```
+  (Evaluation requires including this repository in `ASSET_PATH`.)
+
+### Push-T
+- Train:
+  ```
+  python3 train.py --config-name=train_pusht
+  ```
+  Snapshots will be saved to a new timestamped directory `./exp_local/{date}/{time}_train_pusht_dynamo`.
+  
+  The encoder snapshot will be at `./exp_local/{date}/{time}_train_pusht_dynamo/encoder.pt`
+- In `eval_configs/encoder/pusht_dynamo.yaml`, set `SNAPSHOT_PATH` to the absolute path of the encoder snapshot above.
+- Evaluation:
+  ```
+  python3 online_eval.py --config-name=train_pusht
+  ```
+
+### LIBERO Goal
+- Train:
+  ```
+  python3 train.py --config-name=train_libero_goal
+  ```
+  Snapshots will be saved to a new timestamped directory `./exp_local/{date}/{time}_train_libero_goal_dynamo`.
+
+  The encoder snapshot will be at `./exp_local/{date}/{time}_train_libero_goal_dynamo/encoder.pt`
+- In `eval_configs/encoder/libero_dynamo.yaml`, set `SNAPSHOT_PATH` to the absolute path of the encoder snapshot above.
+- Evaluation:
+  ```
+  MUJOCO_GL=egl python3 online_eval.py --config-name=train_libero_goal
+  ```
+
+## Train on your own dataset
+- Plug in your dataset in these files:
+  - `datasets/your_dataset.py`
+  - `configs/env/your_dataset.yaml`
+  - `configs/env_vars/env_vars.yaml`
+
+- Check the inverse/forward model configs:
+  - `configs/train_your_dataset.yaml`
+    - This is the main config.
+  - `configs/ssl/dynamo_your_dataset.yaml`
+    - If the model converges slowly, try setting `ema_beta` to `null` to use SimSiam instead of EMA encoder during training.
+  - `configs/projector/inverse_dynamics_your_dataset.yaml`
+    - We find that setting the inverse dynamics `output_dim` to approximately the underlying state dimension usually works well.
+      - For sim environments, this is the state-based observation dimension.
+      - For real environments, e.g. a 7DoF robot arm + gripper (1D) manipulating a rigid object (6D), this would be ~16 dimensions.
+
+- Add linear probes for training diagnostics:
+  - `workspaces/your_workspace.py`
+    - This template computes linear probe and nearest neighbor MSE from the image embeddings to states/actions, for monitoring training convergence.
+    - It assumes that your dataset class has `states` (`batch` x `time` x `state_dim`) and `actions` (`batch` x `time` x `action_dim`) attributes.
+      - For a real-world dataset, you can use proprioception as the state.

conda_env.yml

@@ -0,0 +1,204 @@
+name: dynamo-repro
+channels:
+  - conda-forge
+  - defaults
+dependencies:
+  - pip=20.0.2=py38_1
+  - python=3.8.17=he550d4f_0_cpython
+  - readline=8.2=h8228510_1
+  - setuptools=68.0.0=pyhd8ed1ab_0
+  - tk=8.6.12=h27826a3_0
+  - wheel=0.41.1=pyhd8ed1ab_0
+  - xz=5.2.6=h166bdaf_0
+  - pip:
+    - absl-py==1.4.0
+    - accelerate==0.22.0
+    - antlr4-python3-runtime==4.9.3
+    - anyio==3.7.1
+    - appdirs==1.4.4
+    - argon2-cffi==23.1.0
+    - argon2-cffi-bindings==21.2.0
+    - asttokens==2.2.1
+    - async-lru==2.0.4
+    - attrs==23.1.0
+    - av==10.0.0
+    - babel==2.12.1
+    - backcall==0.2.0
+    - bddl==3.5.0
+    - beautifulsoup4==4.12.2
+    - bleach==6.0.0
+    - certifi==2023.7.22
+    - cffi==1.15.1
+    - charset-normalizer==3.2.0
+    - click==8.1.7
+    - cloudpickle==2.2.1
+    - cmake==3.27.2
+    - comm==0.1.4
+    - contourpy==1.1.0
+    - cycler==0.11.0
+    - cython==0.29.37
+    - git+https://github.com/Farama-Foundation/d4rl@71a9549f2091accff93eeff68f1f3ab2c0e0a288#egg=d4rl
+    - debugpy==1.6.7.post1
+    - decorator==4.4.2
+    - decord==0.6.0
+    - defusedxml==0.7.1
+    - dm-control==1.0.14
+    - dm-env==1.6
+    - dm-tree==0.1.8
+    - docker-pycreds==0.4.0
+    - easydict==1.13
+    - einops==0.6.1
+    - evdev==1.6.1
+    - exceptiongroup==1.1.3
+    - executing==1.2.0
+    - fasteners==0.18
+    - fastjsonschema==2.18.0
+    - filelock==3.12.2
+    - fonttools==4.42.1
+    - future==0.18.3
+    - gitdb==4.0.10
+    - gitpython==3.1.32
+    - glfw==2.6.2
+    - gym==0.23.1
+    - gdown==5.1.0
+    - h5py==3.9.0
+    - huggingface-hub==0.22.2
+    - hydra-core==1.3.2
+    - hydra-submitit-launcher==1.2.0
+    - idna==3.4
+    - imageio==2.31.1
+    - imageio-ffmpeg==0.4.8
+    - importlib-resources==6.0.1
+    - iopath==0.1.10
+    - ipdb==0.13.13
+    - ipykernel==6.25.1
+    - ipython==8.12.2
+    - ipywidgets==8.1.0
+    - jedi==0.19.0
+    - jinja2==3.1.2
+    - joblib==1.3.2
+    - json5==0.9.14
+    - jsonschema==4.19.0
+    - jsonschema-specifications==2023.7.1
+    - jupyter-client==8.3.0
+    - jupyter-core==5.3.1
+    - jupyter-events==0.7.0
+    - jupyter-lsp==2.2.0
+    - jupyter-server==2.7.2
+    - jupyter-server-terminals==0.4.4
+    - jupyterlab==4.0.5
+    - jupyterlab-pygments==0.2.2
+    - jupyterlab-server==2.24.0
+    - jupyterlab-widgets==3.0.8
+    - kiwisolver==1.4.4
+    - labmaze==1.0.6
+    - lit==16.0.6
+    - lxml==4.9.3
+    - markupsafe==2.1.3
+    - matplotlib==3.7.2
+    - matplotlib-inline==0.1.6
+    - mistune==3.0.1
+    - moviepy==1.0.3
+    - mpmath==1.3.0
+    - msgpack==1.0.5
+    - mujoco==2.3.7
+    - mujoco-py==2.1.2.14
+    - nbclient==0.8.0
+    - nbconvert==7.7.4
+    - nbformat==5.9.2
+    - nest-asyncio==1.5.7
+    - networkx==3.1
+    - notebook-shim==0.2.3
+    - numpy==1.24.4
+    - nvidia-cublas-cu11==11.10.3.66
+    - nvidia-cuda-cupti-cu11==11.7.101
+    - nvidia-cuda-nvrtc-cu11==11.7.99
+    - nvidia-cuda-runtime-cu11==11.7.99
+    - nvidia-cudnn-cu11==8.5.0.96
+    - nvidia-cufft-cu11==10.9.0.58
+    - nvidia-curand-cu11==10.2.10.91
+    - nvidia-cusolver-cu11==11.4.0.1
+    - nvidia-cusparse-cu11==11.7.4.91
+    - nvidia-nccl-cu11==2.14.3
+    - nvidia-nvtx-cu11==11.7.91
+    - omegaconf==2.3.0
+    - opencv-python==4.8.0.76
+    - overrides==7.4.0
+    - packaging==23.1
+    - pandas==2.0.3
+    - pandocfilters==1.5.0
+    - parso==0.8.3
+    - patchelf==0.17.2.1
+    - pathtools==0.1.2
+    - pexpect==4.8.0
+    - pickleshare==0.7.5
+    - pillow==10.0.0
+    - pkgutil-resolve-name==1.3.10
+    - platformdirs==3.10.0
+    - prettytable==3.8.0
+    - proglog==0.1.10
+    - prometheus-client==0.17.1
+    - prompt-toolkit==3.0.39
+    - protobuf==4.24.1
+    - psutil==5.9.5
+    - ptyprocess==0.7.0
+    - pure-eval==0.2.2
+    - pybullet==3.2.5
+    - pycparser==2.21
+    - pygame==2.5.2
+    - pygments==2.16.1
+    - pymunk==6.6.0
+    - pynput==1.7.6
+    - pynvml==11.5.0
+    - pyopengl==3.1.7
+    - pyopengl-accelerate==3.1.7
+    - pyparsing==3.0.9
+    - python-json-logger==2.0.7
+    - python-xlib==0.33
+    - pyyaml==6.0.1
+    - pyzmq==25.1.0
+    - referencing==0.30.2
+    - requests==2.31.0
+    - rfc3339-validator==0.1.4
+    - rfc3986-validator==0.1.1
+    - robosuite==1.4.1
+    - rpds-py==0.9.2
+    - scikit-image==0.19.3
+    - scikit-learn==1.3.2
+    - scipy==1.10.1
+    - send2trash==1.8.2
+    - sentry-sdk==1.29.2
+    - setproctitle==1.3.2
+    - shapely==2.0.3
+    - six==1.16.0
+    - smmap==5.0.0
+    - sniffio==1.3.0
+    - soupsieve==2.4.1
+    - stack-data==0.6.2
+    - submitit==1.5.1
+    - sympy==1.12
+    - tables==3.8.0
+    - tabulate==0.9.0
+    - termcolor==2.3.0
+    - terminado==0.17.1
+    - threadpoolctl==3.2.0
+    - tifffile==2023.7.10
+    - timm==0.9.16
+    - tinycss2==1.2.1
+    - tomli==2.0.1
+    - torch==2.0.1
+    - torchvision==0.15.2
+    - tornado==6.3.3
+    - tqdm==4.66.1
+    - traitlets==5.9.0
+    - triton==2.0.0
+    - typing-extensions==4.7.1
+    - urllib3==2.0.4
+    - wandb==0.15.8
+    - wcwidth==0.2.6
+    - webencodings==0.5.1
+    - websocket-client==1.6.1
+    - widgetsnbextension==4.0.8
+    - zarr==2.16.1
+    - zipp==3.16.2
+    - git+https://github.com/Farama-Foundation/d4rl.git

configs/encoder/identity.yaml

@@ -0,0 +1,2 @@
+_target_: torch.nn.Identity
+output_dim: ${env.obs_dim}

configs/encoder/resnet18_random.yaml

@@ -0,0 +1,4 @@
+_target_: models.encoder.resnet.resnet18
+pretrained: False
+output_dim: 512
+unit_norm: False

configs/env/block_push_multiview.yaml

@@ -0,0 +1,12 @@
+views: 2
+action_dim: 2
+
+workspace:
+  _target_: workspaces.block_push_multiview.BlockPushMultiviewWorkspace
+
+dataset:
+  _target_: datasets.block_pushing.PushMultiviewTrajectoryDataset
+  data_directory: ${env_vars.datasets.block_push}
+  onehot_goals: False
+  subset_fraction: ${subset_fraction}
+  prefetch: True

configs/env/libero_goal.yaml

@@ -0,0 +1,9 @@
+views: 2
+action_dim: 7
+
+workspace:
+  _target_: workspaces.libero_goal.LiberoGoalWorkspace
+
+dataset:
+  _target_: datasets.libero.LiberoGoalDataset
+  data_directory: ${env_vars.datasets.libero}

configs/env/pusht.yaml

@@ -0,0 +1,11 @@
+views: 1
+action_dim: 2
+
+workspace:
+  _target_: workspaces.pusht.PushTWorkspace
+
+dataset:
+  _target_: datasets.pusht.PushTDataset
+  data_directory: ${env_vars.datasets.pusht}
+  subset_fraction: ${subset_fraction}
+  relative: ${relative}

configs/env/sim_kitchen.yaml

@@ -0,0 +1,11 @@
+views: 1
+action_dim: 9
+
+workspace:
+  _target_: workspaces.sim_kitchen.SimKitchenWorkspace
+
+dataset:
+  _target_: datasets.sim_kitchen.SimKitchenTrajectoryDataset
+  data_directory: ${env_vars.datasets.sim_kitchen}
+  onehot_goals: False
+  prefetch: True

configs/env/your_dataset.yaml

@@ -0,0 +1,8 @@
+views: NUM_VIEWS
+
+workspace:
+  _target_: workspaces.your_workspace.YourWorkspace
+
+dataset:
+  _target_: datasets.your_dataset.YourTrajectoryDataset
+  data_directory: ${env_vars.datasets.your_trajectory_dataset}

configs/env_vars/env_vars.yaml

@@ -0,0 +1,8 @@
+dataset_root: /PATH/TO/DATASET/ROOT  # e.g. set this to the unzipped directory of all datasets
+
+datasets:
+    pusht: ${env_vars.dataset_root}/pusht_dataset
+    sim_kitchen: ${env_vars.dataset_root}/sim_kitchen_dataset
+    libero: ${env_vars.dataset_root}/libero_dataset
+    block_push: ${env_vars.dataset_root}/block_push_dataset
+    your_trajectory_dataset: YOUR_DATASET_PATH

configs/projector/inverse_dynamics_blockpush.yaml

@@ -0,0 +1,8 @@
+_target_: models.projector.inverse_dynamics.InverseDynamicsProjector
+window_size: ${window_size}
+input_dim: ${encoder.output_dim}
+n_layer: 4
+n_head: 4
+n_embd: 72
+output_dim: 16
+dropout: 0.0

configs/projector/inverse_dynamics_libero.yaml

@@ -0,0 +1,8 @@
+_target_: models.projector.inverse_dynamics.InverseDynamicsProjector
+window_size: ${window_size}
+input_dim: ${encoder.output_dim}
+n_layer: 6
+n_head: 6
+n_embd: 120
+output_dim: 32
+dropout: 0.0

configs/projector/inverse_dynamics_pusht.yaml

@@ -0,0 +1,8 @@
+_target_: models.projector.inverse_dynamics.InverseDynamicsProjector
+window_size: ${window_size}
+input_dim: ${encoder.output_dim}
+n_layer: 6
+n_head: 6
+n_embd: 120
+output_dim: 8
+dropout: 0.0

configs/projector/inverse_dynamics_sim_kitchen.yaml

@@ -0,0 +1,8 @@
+_target_: models.projector.inverse_dynamics.InverseDynamicsProjector
+window_size: ${window_size}
+input_dim: ${encoder.output_dim}
+n_layer: 6
+n_head: 6
+n_embd: 120
+output_dim: 64
+dropout: 0.0

configs/projector/inverse_dynamics_your_dataset.yaml

@@ -0,0 +1,11 @@
+_target_: models.projector.inverse_dynamics.InverseDynamicsProjector
+window_size: ${window_size}
+input_dim: ${encoder.output_dim}
+n_layer: 6
+n_head: 6
+n_embd: 120
+# output_dim: for sim environments, set it to the state-based observation dimension.
+# for real environments, set it to the estimated underlying environment state dimension.
+# (e.g. if we have a 7DoF robot arm and a free rigid object, 16 should work fine)
+output_dim: OUTPUT_DIM
+dropout: 0.0

configs/ssl/dynamo_blockpush.yaml

@@ -0,0 +1,20 @@
+_target_: models.ssl.dynamo.DynaMoSSL
+window_size: ${window_size}
+feature_dim: ${encoder.output_dim}
+projection_dim: ${projector.output_dim}
+n_layer: 4
+n_head: 4
+n_embd: 72
+
+dropout: 0.3 # dropout on the forward dynamics model
+covariance_reg_coef: 0.04
+dynamics_loss_coef: 1.0
+
+ema_beta: 0.99
+beta_scheduling: True
+projector_use_ema: True
+
+lr: ${ssl_lr}
+weight_decay: ${ssl_weight_decay}
+betas: ${betas}
+separate_single_views: True

configs/ssl/dynamo_libero.yaml

@@ -0,0 +1,20 @@
+_target_: models.ssl.dynamo.DynaMoSSL
+window_size: ${window_size}
+feature_dim: ${encoder.output_dim}
+projection_dim: ${projector.output_dim}
+n_layer: 6
+n_head: 6
+n_embd: 120
+
+dropout: 0.0 # dropout on the forward dynamics model
+covariance_reg_coef: 0.04
+dynamics_loss_coef: 1.0
+
+ema_beta: null
+beta_scheduling: True
+projector_use_ema: True
+
+lr: ${ssl_lr}
+weight_decay: ${ssl_weight_decay}
+betas: ${betas}
+separate_single_views: True

configs/ssl/dynamo_pusht.yaml

@@ -0,0 +1,20 @@
+_target_: models.ssl.dynamo.DynaMoSSL
+window_size: ${window_size}
+feature_dim: ${encoder.output_dim}
+projection_dim: ${projector.output_dim}
+n_layer: 6
+n_head: 6
+n_embd: 120
+
+dropout: 0.0 # dropout on the forward dynamics model
+covariance_reg_coef: 0.04
+dynamics_loss_coef: 1.0
+
+ema_beta: null
+beta_scheduling: True
+projector_use_ema: True
+
+lr: ${ssl_lr}
+weight_decay: ${ssl_weight_decay}
+betas: ${betas}
+separate_single_views: True

configs/ssl/dynamo_sim_kitchen.yaml

@@ -0,0 +1,20 @@
+_target_: models.ssl.dynamo.DynaMoSSL
+window_size: ${window_size}
+feature_dim: ${encoder.output_dim}
+projection_dim: ${projector.output_dim}
+n_layer: 6
+n_head: 6
+n_embd: 120
+
+dropout: 0.0 # dropout on the forward dynamics model
+covariance_reg_coef: 0.04
+dynamics_loss_coef: 1.0
+
+ema_beta: null
+beta_scheduling: True
+projector_use_ema: True
+
+lr: ${ssl_lr}
+weight_decay: ${ssl_weight_decay}
+betas: ${betas}
+separate_single_views: True

configs/ssl/dynamo_your_dataset.yaml

@@ -0,0 +1,20 @@
+_target_: models.ssl.dynamo.DynaMoSSL
+window_size: ${window_size}
+feature_dim: ${encoder.output_dim}
+projection_dim: ${projector.output_dim}
+n_layer: 6
+n_head: 6
+n_embd: 120
+
+dropout: 0.0 # dropout on the forward dynamics model
+covariance_reg_coef: 0.04
+dynamics_loss_coef: 1.0
+
+ema_beta: 0.99  # set to null for SimSiam instead of EMA
+beta_scheduling: True
+projector_use_ema: True
+
+lr: ${ssl_lr}
+weight_decay: ${ssl_weight_decay}
+betas: ${betas}
+separate_single_views: True

configs/train_blockpush.yaml

@@ -0,0 +1,53 @@
+defaults:
+  - _self_
+  - encoder: resnet18_random
+  - projector: inverse_dynamics_blockpush
+  - ssl: dynamo_blockpush
+  - env: block_push_multiview
+  - env_vars: env_vars
+
+# Dataset details
+subset_fraction: null
+train_fraction: 0.95
+batch_size: 64  # across all processes
+num_workers: 15  # per process
+window_size: 5
+goal_conditional: null
+goal_seq_len: 0
+min_future_sep: 0
+num_extra_predicted_actions: 0
+
+# Training details
+ssl_lr: 1e-4
+ssl_weight_decay: 0.0
+betas:
+  - 0.9
+  - 0.999
+clip_grad_norm: 0.1
+seed: 42
+timeout_seconds: 18000
+
+sync_bn: True
+use_lr_scheduling: True
+warmup_epochs: 5
+num_epochs: 40
+
+save_every_epochs: 10
+
+# Eval config
+eval_offline: True
+
+# Wandb config
+project: dynamo-repro
+experiment: train_blockpush_dynamo
+
+# hydra config
+hydra:
+  job:
+    override_dirname: ${experiment}
+    chdir: False
+  run:
+    dir: ./exp_local/${now:%Y.%m.%d}/${now:%H%M%S}_${experiment}
+  sweep:
+    dir: ./exp_local/${now:%Y.%m.%d}/sweep_${now:%H%M%S}_${experiment}
+    subdir: ${hydra.job.num}

configs/train_libero_goal.yaml

@@ -0,0 +1,53 @@
+defaults:
+  - _self_
+  - encoder: resnet18_random
+  - projector: inverse_dynamics_libero
+  - ssl: dynamo_libero
+  - env: libero_goal
+  - env_vars: env_vars
+
+# Dataset details
+subset_fraction: null
+train_fraction: 0.95
+batch_size: 64  # across all processes
+num_workers: 15  # per process
+window_size: 5
+goal_conditional: null
+goal_seq_len: 0
+min_future_sep: 0
+num_extra_predicted_actions: 0
+
+# Training details
+ssl_lr: 1e-4
+ssl_weight_decay: 0.0
+betas:
+  - 0.9
+  - 0.999
+clip_grad_norm: 0.1
+seed: 42
+timeout_seconds: 18000
+
+sync_bn: True
+use_lr_scheduling: False
+warmup_epochs: 5
+num_epochs: 40
+
+save_every_epochs: 10
+
+# Eval config
+eval_offline: True
+
+# Wandb config
+project: dynamo-repro
+experiment: train_libero_goal_dynamo
+
+# hydra config
+hydra:
+  job:
+    override_dirname: ${experiment}
+    chdir: False
+  run:
+    dir: ./exp_local/${now:%Y.%m.%d}/${now:%H%M%S}_${experiment}
+  sweep:
+    dir: ./exp_local/${now:%Y.%m.%d}/sweep_${now:%H%M%S}_${experiment}
+    subdir: ${hydra.job.num}

configs/train_pusht.yaml

@@ -0,0 +1,54 @@
+defaults:
+  - _self_
+  - encoder: resnet18_random
+  - projector: inverse_dynamics_pusht
+  - ssl: dynamo_pusht
+  - env: pusht
+  - env_vars: env_vars
+
+# Dataset details
+subset_fraction: null
+train_fraction: 0.95
+batch_size: 64
+num_workers: 15
+window_size: 5
+goal_conditional: null
+goal_seq_len: 0
+min_future_sep: 0
+num_extra_predicted_actions: 5
+relative: False
+
+# Training details
+ssl_lr: 1e-4
+ssl_weight_decay: 1e-6
+betas:
+  - 0.9
+  - 0.999
+clip_grad_norm: 0.1
+seed: 42
+timeout_seconds: 18000
+
+sync_bn: True
+use_lr_scheduling: True
+warmup_epochs: 5
+num_epochs: 40
+
+save_every_epochs: 10
+
+# Eval config
+eval_offline: True
+
+# Wandb config
+project: dynamo-repro
+experiment: train_pusht_dynamo
+
+# hydra config
+hydra:
+  job:
+    override_dirname: ${experiment}
+    chdir: False
+  run:
+    dir: ./exp_local/${now:%Y.%m.%d}/${now:%H%M%S}_${experiment}
+  sweep:
+    dir: ./exp_local/${now:%Y.%m.%d}/sweep_${now:%H%M%S}_${experiment}
+    subdir: ${hydra.job.num}

configs/train_sim_kitchen.yaml

@@ -0,0 +1,53 @@
+defaults:
+  - _self_
+  - encoder: resnet18_random
+  - projector: inverse_dynamics_sim_kitchen
+  - ssl: dynamo_sim_kitchen
+  - env: sim_kitchen
+  - env_vars: env_vars
+
+# Dataset details
+subset_fraction: null
+train_fraction: 0.95
+batch_size: 64  # across all processes
+num_workers: 15  # per process
+window_size: 2
+goal_conditional: null
+goal_seq_len: 3
+min_future_sep: 10
+num_extra_predicted_actions: 0
+
+# Training details
+ssl_lr: 1e-4
+ssl_weight_decay: 0.0
+betas:
+  - 0.9
+  - 0.999
+clip_grad_norm: 0.1
+seed: 42
+timeout_seconds: 18000
+
+sync_bn: True
+use_lr_scheduling: True
+warmup_epochs: 5
+num_epochs: 40
+
+save_every_epochs: 10
+
+# Eval config
+eval_offline: True
+
+# Wandb config
+project: dynamo-repro
+experiment: train_sim_kitchen_dynamo
+
+# hydra config
+hydra:
+  job:
+    override_dirname: ${experiment}
+    chdir: False
+  run:
+    dir: ./exp_local/${now:%Y.%m.%d}/${now:%H%M%S}_${experiment}
+  sweep:
+    dir: ./exp_local/${now:%Y.%m.%d}/sweep_${now:%H%M%S}_${experiment}
+    subdir: ${hydra.job.num}

configs/train_your_dataset.yaml

@@ -0,0 +1,53 @@
+defaults:
+  - _self_
+  - encoder: resnet18_random
+  - projector: inverse_dynamics_your_dataset
+  - ssl: dynamo_your_dataset
+  - env: your_dataset
+  - env_vars: env_vars
+
+# Dataset details
+subset_fraction: null
+train_fraction: 0.95
+batch_size: 64  # across all processes
+num_workers: 15  # per process
+window_size: 5
+goal_conditional: null
+goal_seq_len: 0
+min_future_sep: 0
+num_extra_predicted_actions: 0
+
+# Training details
+ssl_lr: 1e-4
+ssl_weight_decay: 0.0
+betas:
+  - 0.9
+  - 0.999
+clip_grad_norm: 0.1
+seed: 42
+timeout_seconds: 18000
+
+sync_bn: True
+use_lr_scheduling: True
+warmup_epochs: 5
+num_epochs: 40
+
+save_every_epochs: 10
+
+# Eval config
+eval_offline: True
+
+# Wandb config
+project: dynamo-repro
+experiment: train_your_dataset_dynamo
+
+# hydra config
+hydra:
+  job:
+    override_dirname: ${experiment}
+    chdir: False
+  run:
+    dir: ./exp_local/${now:%Y.%m.%d}/${now:%H%M%S}_${experiment}
+  sweep:
+    dir: ./exp_local/${now:%Y.%m.%d}/sweep_${now:%H%M%S}_${experiment}
+    subdir: ${hydra.job.num}

datasets/__init__.py

@@ -0,0 +1,5 @@
+from . import core
+from . import block_pushing
+from . import libero
+from . import sim_kitchen
+from . import pusht

datasets/block_pushing.py

@@ -0,0 +1,79 @@
+import os
+import torch
+import einops
+import numpy as np
+from pathlib import Path
+from typing import Optional
+from datasets.core import TrajectoryDataset
+
+
+class PushMultiviewTrajectoryDataset(TrajectoryDataset):
+    def __init__(
+        self,
+        data_directory: os.PathLike,
+        onehot_goals=False,
+        subset_fraction: Optional[float] = None,
+        prefetch: bool = False,
+    ):
+        self.data_directory = Path(data_directory)
+        self.states = np.load(self.data_directory / "multimodal_push_observations.npy")
+        self.actions = np.load(self.data_directory / "multimodal_push_actions.npy")
+        self.masks = np.load(self.data_directory / "multimodal_push_masks.npy")
+
+        self.subset_fraction = subset_fraction
+        if self.subset_fraction:
+            assert self.subset_fraction > 0 and self.subset_fraction <= 1
+            n = int(len(self.states) * self.subset_fraction)
+        else:
+            n = len(self.states)
+        self.states = self.states[:n]
+        self.actions = self.actions[:n]
+        self.masks = self.masks[:n]
+
+        self.states = torch.from_numpy(self.states).float()
+        self.actions = torch.from_numpy(self.actions).float() / 0.03
+        self.masks = torch.from_numpy(self.masks).bool()
+        self.prefetch = prefetch
+        if self.prefetch:
+            self.obses = []
+            for i in range(n):
+                vid_path = self.data_directory / "obs_multiview" / f"{i:03d}.pth"
+                self.obses.append(torch.load(vid_path))
+        self.onehot_goals = onehot_goals
+        if self.onehot_goals:
+            self.goals = torch.load(self.data_directory / "onehot_goals.pth").float()
+            self.goals = self.goals[:n]
+
+    def get_seq_length(self, idx):
+        return int(self.masks[idx].sum().item())
+
+    def get_all_actions(self):
+        result = []
+        # mask out invalid actions
+        for i in range(len(self.masks)):
+            T = int(self.masks[i].sum().item())
+            result.append(self.actions[i, :T, :])
+        return torch.cat(result, dim=0)
+
+    def get_frames(self, idx, frames):
+        if self.prefetch:
+            obs = self.obses[idx][frames]
+        else:
+            obs = torch.load(self.data_directory / "obs_multiview" / f"{idx:03d}.pth")[
+                frames
+            ]
+        obs = einops.rearrange(obs, "T V H W C -> T V C H W") / 255.0
+        act = self.actions[idx, frames]
+        mask = self.masks[idx, frames]
+        if self.onehot_goals:
+            goal = self.goals[idx, frames]
+            return obs, act, mask, goal
+        else:
+            return obs, act, mask
+
+    def __getitem__(self, idx):
+        T = self.masks[idx].sum().int().item()
+        return self.get_frames(idx, range(T))
+
+    def __len__(self):
+        return len(self.states)

datasets/core.py

@@ -0,0 +1,345 @@
+import abc
+import utils
+import torch
+import numpy as np
+from torch import default_generator, randperm
+from torch.utils.data import Dataset, Subset
+from typing import Callable, Optional, Sequence, List, Any
+from torch.nn.utils.rnn import pad_sequence
+
+
+# Taken from python 3.5 docs
+def _accumulate(iterable, fn=lambda x, y: x + y):
+    "Return running totals"
+    # _accumulate([1,2,3,4,5]) --> 1 3 6 10 15
+    # _accumulate([1,2,3,4,5], operator.mul) --> 1 2 6 24 120
+    it = iter(iterable)
+    try:
+        total = next(it)
+    except StopIteration:
+        return
+    yield total
+    for element in it:
+        total = fn(total, element)
+        yield total
+
+
+class TrajectoryDataset(Dataset, abc.ABC):
+    """
+    A dataset containing trajectories.
+    TrajectoryDataset[i] returns: (observations, actions, mask)
+        observations: Tensor[T, ...], T frames of observations
+        actions: Tensor[T, ...], T frames of actions
+        mask: Tensor[T]: False: invalid; True: valid
+    """
+
+    @abc.abstractmethod
+    def get_seq_length(self, idx):
+        """
+        Returns the length of the idx-th trajectory.
+        """
+        raise NotImplementedError
+
+    @abc.abstractmethod
+    def get_frames(self, idx, frames):
+        """
+        Returns the frames from the idx-th trajectory at the specified frames.
+        Used to speed up slicing.
+        """
+        raise NotImplementedError
+
+
+class TrajectorySubset(TrajectoryDataset, Subset):
+    """
+    Subset of a trajectory dataset at specified indices.
+
+    Args:
+        dataset (TrajectoryDataset): The whole Dataset
+        indices (sequence): Indices in the whole set selected for subset
+    """
+
+    def __init__(self, dataset: TrajectoryDataset, indices: Sequence[int]):
+        Subset.__init__(self, dataset, indices)
+
+    def get_seq_length(self, idx):
+        return self.dataset.get_seq_length(self.indices[idx])
+
+    def get_all_actions(self):
+        return self.dataset.get_all_actions()
+
+    def get_frames(self, idx, frames):
+        return self.dataset.get_frames(self.indices[idx], frames)
+
+
+class TrajectorySlicerDataset:
+    def __init__(
+        self,
+        dataset: TrajectoryDataset,
+        window: int,
+        future_conditional: bool = False,
+        min_future_sep: int = 0,
+        future_seq_len: Optional[int] = None,
+        only_sample_tail: bool = False,
+        transform: Optional[Callable] = None,
+        num_extra_predicted_actions: Optional[int] = None,
+        frame_step: int = 1,
+        repeat_first_frame: bool = False,
+    ):
+        """
+        Slice a trajectory dataset into unique (but overlapping) sequences of length `window`.
+
+        dataset: a trajectory dataset that satisfies:
+            dataset.get_seq_length(i) is implemented to return the length of sequence i
+            dataset[i] = (observations, actions, mask)
+            observations: Tensor[T, ...]
+            actions: Tensor[T, ...]
+            mask: Tensor[T]
+                False: invalid
+                True: valid
+        window: int
+            number of timesteps to include in each slice
+        future_conditional: bool = False
+            if True, observations will be augmented with future observations sampled from the same trajectory
+        min_future_sep: int = 0
+            minimum number of timesteps between the end of the current sequence and the start of the future sequence
+            for the future conditional
+        future_seq_len: Optional[int] = None
+            the length of the future conditional sequence;
+            required if future_conditional is True
+        only_sample_tail: bool = False
+            if True, only sample future sequences from the tail of the trajectory
+        transform: function (observations, actions, mask[, goal]) -> (observations, actions, mask[, goal])
+        """
+        if future_conditional:
+            assert future_seq_len is not None, "must specify a future_seq_len"
+        self.dataset = dataset
+        self.window = window
+        self.future_conditional = future_conditional
+        self.min_future_sep = min_future_sep
+        self.future_seq_len = future_seq_len
+        self.only_sample_tail = only_sample_tail
+        self.transform = transform
+        self.num_extra_predicted_actions = num_extra_predicted_actions or 0
+        self.slices = []
+        self.frame_step = frame_step
+        min_seq_length = np.inf
+        if num_extra_predicted_actions:
+            window = window + num_extra_predicted_actions
+        for i in range(len(self.dataset)):  # type: ignore
+            T = self.dataset.get_seq_length(i)  # avoid reading actual seq (slow)
+            min_seq_length = min(T, min_seq_length)
+            if T - window < 0:
+                print(f"Ignored short sequence #{i}: len={T}, window={window}")
+            else:
+                if repeat_first_frame:
+                    self.slices += [(i, 0, end + 1) for end in range(window - 1)]
+                window_len_with_step = (window - 1) * frame_step + 1
+                last_start = T - window_len_with_step
+                self.slices += [
+                    (i, start, start + window_len_with_step)
+                    for start in range(last_start)
+                ]  # slice indices follow convention [start, end)
+
+        if min_seq_length < window:
+            print(
+                f"Ignored short sequences. To include all, set window <= {min_seq_length}."
+            )
+
+    def get_seq_length(self, idx: int) -> int:
+        if self.future_conditional:
+            return self.future_seq_len + self.window
+        else:
+            return self.window
+
+    def get_all_actions(self) -> torch.Tensor:
+        return self.dataset.get_all_actions()
+
+    def __len__(self):
+        return len(self.slices)
+
+    def __getitem__(self, idx):
+        i, start, end = self.slices[idx]
+        T = self.dataset.get_seq_length(i)
+
+        if (
+            self.num_extra_predicted_actions is not None
+            and self.num_extra_predicted_actions != 0
+        ):
+            assert self.frame_step == 1, "NOT TESTED"
+            if self.future_conditional:
+                raise NotImplementedError(
+                    "num_extra_predicted_actions with future_conditional not implemented"
+                )
+            assert end <= T, f"end={end} > T={T}"
+            observations, actions, mask = self.dataset.get_frames(i, range(start, end))
+            observations = observations[: self.window]
+
+            values = [observations, actions, mask.bool()]
+        else:
+            if self.future_conditional:
+                assert self.frame_step == 1, "NOT TESTED"
+                valid_start_range = (
+                    end + self.min_future_sep,
+                    self.dataset.get_seq_length(i) - self.future_seq_len,
+                )
+                if valid_start_range[0] < valid_start_range[1]:
+                    if self.only_sample_tail:
+                        future_obs_range = range(T - self.future_seq_len, T)
+                    else:
+                        future_start = np.random.randint(*valid_start_range)
+                        future_end = future_start + self.future_seq_len
+                        future_obs_range = range(future_start, future_end)
+                    obs, actions, mask = self.dataset.get_frames(
+                        i, list(range(start, end)) + list(future_obs_range)
+                    )
+                    future_obs = obs[end - start :]
+                    obs = obs[: end - start]
+                    actions = actions[: end - start]
+                    mask = mask[: end - start]
+                else:
+                    # zeros placeholder T x obs_dim
+                    obs, actions, mask = self.dataset.get_frames(i, range(start, end))
+                    obs_dims = obs.shape[1:]
+                    future_obs = torch.zeros((self.future_seq_len, *obs_dims))
+
+                # [observations, actions, mask, future_obs (goal conditional)]
+                values = [obs, actions, mask.bool(), future_obs]
+            else:
+                observations, actions, mask = self.dataset.get_frames(
+                    i, range(start, end, self.frame_step)
+                )
+                values = [observations, actions, mask.bool()]
+
+        if end - start < self.window + self.num_extra_predicted_actions:
+            # this only happens for repeating the very first frames
+            values = [
+                utils.inference.repeat_start_to_length(
+                    x, self.window + self.num_extra_predicted_actions, dim=0
+                )
+                for x in values
+            ]
+            values[0] = values[0][: self.window]
+
+        # optionally apply transform
+        if self.transform is not None:
+            values = self.transform(values)
+        return tuple(values)
+
+
+class TrajectoryEmbeddingDataset(TrajectoryDataset):
+    def __init__(
+        self,
+        model,
+        dataset: TrajectoryDataset,
+        device="cpu",
+        embed_goal=False,
+    ):
+        self.data = utils.inference.embed_trajectory_dataset(
+            model,
+            dataset,
+            obs_only=False,
+            device=device,
+            embed_goal=embed_goal,
+        )
+        assert len(self.data) == len(dataset)
+
+        self.seq_lengths = [len(x[0]) for x in self.data]
+        self.on_device_data = []
+        n_tensors = len(self.data[0])
+        for i in range(n_tensors):
+            self.on_device_data.append(
+                pad_sequence([x[i] for x in self.data], batch_first=True).to(device)
+            )
+        self.data = self.on_device_data
+
+    def get_seq_length(self, idx):
+        return self.seq_lengths[idx]
+
+    def get_all_actions(self):
+        return torch.cat([x[1] for x in self.data], dim=0)
+
+    def get_frames(self, idx, frames):
+        return [x[idx, frames] for x in self.data]
+
+    def __getitem__(self, idx):
+        return self.get_frames(idx, range(self.get_seq_length(idx)))
+
+    def __len__(self):
+        return len(self.seq_lengths)
+
+
+def get_train_val_sliced(
+    traj_dataset: TrajectoryDataset,
+    train_fraction: float = 0.9,
+    random_seed: int = 42,
+    window_size: int = 10,
+    future_conditional: bool = False,
+    min_future_sep: int = 0,
+    future_seq_len: Optional[int] = None,
+    only_sample_tail: bool = False,
+    transform: Optional[Callable[[Any], Any]] = None,
+    num_extra_predicted_actions: Optional[int] = None,
+    frame_step: int = 1,
+):
+    train, val = split_traj_datasets(
+        traj_dataset,
+        train_fraction=train_fraction,
+        random_seed=random_seed,
+    )
+    traj_slicer_kwargs = {
+        "window": window_size,
+        "future_conditional": future_conditional,
+        "min_future_sep": min_future_sep,
+        "future_seq_len": future_seq_len,
+        "only_sample_tail": only_sample_tail,
+        "transform": transform,
+        "num_extra_predicted_actions": num_extra_predicted_actions,
+        "frame_step": frame_step,
+    }
+
+    train_slices = TrajectorySlicerDataset(train, **traj_slicer_kwargs)
+    val_slices = TrajectorySlicerDataset(val, **traj_slicer_kwargs)
+    return train_slices, val_slices
+
+
+def random_split_traj(
+    dataset: TrajectoryDataset,
+    lengths: Sequence[int],
+    generator: Optional[torch.Generator] = default_generator,
+) -> List[TrajectorySubset]:
+    """
+    (Modified from torch.utils.data.dataset.random_split)
+
+    Randomly split a trajectory dataset into non-overlapping new datasets of given lengths.
+    Optionally fix the generator for reproducible results, e.g.:
+
+    >>> random_split_traj(range(10), [3, 7], generator=torch.Generator().manual_seed(42))
+
+    Args:
+        dataset (TrajectoryDataset): TrajectoryDataset to be split
+        lengths (sequence): lengths of splits to be produced
+        generator (Generator): Generator used for the random permutation.
+    """
+    # Cannot verify that dataset is Sized
+    if sum(lengths) != len(dataset):  # type: ignore[arg-type]
+        raise ValueError(
+            "Sum of input lengths does not equal the length of the input dataset!"
+        )
+
+    indices = randperm(sum(lengths), generator=generator).tolist()
+    return [
+        TrajectorySubset(dataset, indices[offset - length : offset])
+        for offset, length in zip(_accumulate(lengths), lengths)
+    ]
+
+
+def split_traj_datasets(dataset, train_fraction=0.95, random_seed=42):
+    dataset_length = len(dataset)
+    lengths = [
+        int(train_fraction * dataset_length),
+        dataset_length - int(train_fraction * dataset_length),
+    ]
+    train_set, val_set = random_split_traj(
+        dataset, lengths, generator=torch.Generator().manual_seed(random_seed)
+    )
+    return train_set, val_set

datasets/libero.py

@@ -0,0 +1,120 @@
+import torch
+import einops
+import numpy as np
+from pathlib import Path
+from typing import Optional
+from torch.nn.utils.rnn import pad_sequence
+from datasets.core import TrajectoryDataset
+
+
+class LiberoGoalDataset(TrajectoryDataset):
+    # data structure:
+    # libero_goal
+    #      task_name
+    #          demo_{i}
+    #              agentview_image.mp4
+    #              robot0_eye_in_hand_image.mp4
+    #              robot0_joint_pos.npy
+    #              robot0_eef.npy
+    #              robot0_gripper_qpos.npy
+    #              object_states.npy
+    #              actions.npy
+    def __init__(self, data_directory, subset_fraction: Optional[float] = None):
+        self.dir = Path(data_directory) / "libero_goal"
+        self.task_names = list(self.dir.iterdir())
+        self.task_names.sort()
+        self.demos = []
+        for task_name in self.task_names:
+            self.demos += list(task_name.iterdir())
+
+        self.subset_fraction = subset_fraction
+        if self.subset_fraction:
+            assert 0 < self.subset_fraction <= 1
+            n = int(len(self.demos) * self.subset_fraction)
+            self.demos = self.demos[:n]
+
+        # prefetch all npy data
+        self.joint_pos = []
+        self.eef = []
+        self.gripper_qpos = []
+        self.object_states = []
+        self.states = []
+        self.actions = []
+        for demo in self.demos:
+            self.joint_pos.append(np.load(demo / "robot0_joint_pos.npy"))
+            self.eef.append(np.load(demo / "robot0_eef.npy"))
+            self.gripper_qpos.append(np.load(demo / "robot0_gripper_pos.npy"))
+            self.object_states.append(np.load(demo / "object_states.npy"))
+            state = np.concatenate(
+                [
+                    self.joint_pos[-1],
+                    self.eef[-1],
+                    self.gripper_qpos[-1],
+                    self.object_states[-1],
+                ],
+                axis=1,
+            )
+            act = np.load(demo / "actions.npy")
+            self.states.append(torch.from_numpy(state))
+            self.actions.append(torch.from_numpy(act))
+
+        # pad state dimension to same length for linear probe diagnostics
+        MAX_DIM = 128
+        for i in range(len(self.states)):
+            self.states[i] = torch.cat(
+                [
+                    self.states[i],
+                    torch.zeros(
+                        self.states[i].shape[0], MAX_DIM - self.states[i].shape[1]
+                    ),
+                ],
+                dim=1,
+            )
+        # pad states and actions to the same time length
+        self.states = pad_sequence(self.states, batch_first=True).float()
+        self.actions = pad_sequence(self.actions, batch_first=True).float()
+
+        # last frame goal
+        self.goals = None
+        goals = []
+        for i in range(0, 500, 50):
+            last_obs, _, _ = self.get_frames(i, [-1])  # 1 V C H W
+            goals.append(last_obs)
+        self.goals = goals
+
+    def __len__(self):
+        return len(self.demos)
+
+    def get_frames(self, idx, frames):
+        demo = self.demos[idx]
+        agentview_obs = torch.load(
+            str(demo / "agentview_image.pth"),
+        )
+        robotview_obs = torch.load(
+            str(demo / "robot0_eye_in_hand_image.pth"),
+        )
+        agentview = agentview_obs[frames]
+        robotview = robotview_obs[frames]
+        obs = torch.stack([agentview, robotview], dim=1)
+        obs = einops.rearrange(obs, "T V H W C -> T V C H W") / 255.0
+        act = self.actions[idx][frames]
+
+        if self.goals is not None:
+            task_idx = idx // 50
+            goal = self.goals[task_idx].repeat(len(frames), 1, 1, 1, 1)
+            return obs, act, goal
+        else:
+            return obs, act, None
+
+    def __getitem__(self, idx):
+        return self.get_frames(idx, range(len(self.joint_pos[idx])))
+
+    def get_seq_length(self, idx):
+        return len(self.joint_pos[idx])
+
+    def get_all_actions(self):
+        actions = []
+        for i in range(len(self.demos)):
+            T = self.get_seq_length(i)
+            actions.append(self.actions[i][:T])
+        return torch.cat(actions, dim=0)

datasets/pusht.py

@@ -0,0 +1,63 @@
+import torch
+import einops
+import pickle
+from pathlib import Path
+from typing import Optional
+from datasets.core import TrajectoryDataset
+
+
+class PushTDataset(TrajectoryDataset):
+    def __init__(
+        self,
+        data_directory,
+        subset_fraction: Optional[float] = None,
+        relative=False,
+    ):
+        self.data_directory = Path(data_directory)
+        self.relative = relative
+        self.states = torch.load(self.data_directory / "states.pth")
+        if relative:
+            self.actions = torch.load(self.data_directory / "rel_actions.pth")
+        else:
+            self.actions = torch.load(self.data_directory / "abs_actions.pth")
+        with open(self.data_directory / "seq_lengths.pkl", "rb") as f:
+            self.seq_lengths = pickle.load(f)
+
+        self.subset_fraction = subset_fraction
+        if self.subset_fraction:
+            assert self.subset_fraction > 0 and self.subset_fraction <= 1
+            n = int(len(self.states) * self.subset_fraction)
+        else:
+            n = len(self.states)
+        self.states = self.states[:n]
+        self.actions = self.actions[:n]
+        self.seq_lengths = self.seq_lengths[:n]
+
+        for i in range(n):
+            T = self.seq_lengths[i]
+            self.actions[i, T:] = 0  # redo zero padding
+
+    def get_seq_length(self, idx):
+        return self.seq_lengths[idx]
+
+    def get_all_actions(self):
+        result = []
+        for i in range(len(self.seq_lengths)):
+            T = self.seq_lengths[i]
+            result.append(self.actions[i, :T, :])
+        return torch.cat(result, dim=0)
+
+    def get_frames(self, idx, frames):
+        vid_dir = self.data_directory / "obses"
+        obs = torch.load(str(vid_dir / f"episode_{idx:03d}.pth"))
+        obs = obs[frames]  # THWC
+        obs = einops.rearrange(obs, "T H W C -> T 1 C H W") / 255.0  # T V C H W, 1 view
+        act = self.actions[idx, frames]
+        mask = torch.ones(len(act)).bool()
+        return obs, act, mask
+
+    def __getitem__(self, idx):
+        return self.get_frames(idx, range(self.get_seq_length(idx)))
+
+    def __len__(self):
+        return len(self.seq_lengths)

datasets/sim_kitchen.py

@@ -0,0 +1,58 @@
+import utils
+import torch
+import numpy as np
+from pathlib import Path
+from datasets.core import TrajectoryDataset
+
+
+class SimKitchenTrajectoryDataset(TrajectoryDataset):
+    def __init__(self, data_directory, prefetch=True, onehot_goals=False):
+        self.data_directory = Path(data_directory)
+        states = torch.from_numpy(np.load(self.data_directory / "observations_seq.npy"))
+        actions = torch.from_numpy(np.load(self.data_directory / "actions_seq.npy"))
+        goals = torch.load(self.data_directory / "onehot_goals.pth")
+        # The current values are in shape T x N x Dim, move to N x T x Dim
+        self.states, self.actions, self.goals = utils.transpose_batch_timestep(
+            states, actions, goals
+        )
+        self.Ts = np.load(self.data_directory / "existence_mask.npy").sum(axis=0).astype(int).tolist()
+        
+        self.prefetch = prefetch
+        if self.prefetch:
+            self.obses = []
+            for i in range(len(self.Ts)):
+                self.obses.append(torch.load(self.data_directory / "obses" / f"{i:03d}.pth"))
+        self.onehot_goals = onehot_goals
+
+    def get_seq_length(self, idx):
+        return self.Ts[idx]
+
+    def get_all_actions(self):
+        result = []
+        # mask out invalid actions
+        for i in range(len(self.Ts)):
+            T = self.Ts[i]
+            result.append(self.actions[i, :T, :])
+        return torch.cat(result, dim=0)
+
+    def get_frames(self, idx, frames):
+        # obs, act, mask / obs, act, mask, goal
+        if self.prefetch:
+            obs = self.obses[idx][frames]
+        else:
+            obs = torch.load(self.data_directory / "obses" / f"{idx:03d}.pth")[frames]
+        obs = obs / 255.0
+        act = self.actions[idx, frames]
+        mask = torch.ones((len(frames)))
+        if self.onehot_goals:
+            goal = self.goals[idx, frames]
+            return obs, act, mask, goal
+        else:
+            return obs, act, mask
+
+    def __getitem__(self, idx):
+        T = self.Ts[idx]
+        return self.get_frames(idx, range(T))
+    
+    def __len__(self):
+        return len(self.Ts)

datasets/vqbet_repro.py

@@ -0,0 +1,120 @@
+import abc
+import utils
+import torch
+import numpy as np
+from torch.utils.data import Dataset
+from typing import Optional, Callable
+
+
+class TrajectoryDataset(Dataset, abc.ABC):
+    """
+    A dataset containing trajectories.
+    TrajectoryDataset[i] returns: (observations, actions, mask)
+        observations: Tensor[T, ...], T frames of observations
+        actions: Tensor[T, ...], T frames of actions
+        mask: Tensor[T]: 0: invalid; 1: valid
+    """
+
+    @abc.abstractmethod
+    def get_seq_length(self, idx):
+        """
+        Returns the length of the idx-th trajectory.
+        """
+        raise NotImplementedError
+
+
+class TrajectorySlicerDataset(TrajectoryDataset):
+    def __init__(
+        self,
+        dataset: TrajectoryDataset,
+        window: int,
+        action_window: int,
+        vqbet_get_future_action_chunk: bool = True,
+        future_conditional: bool = False,
+        min_future_sep: int = 0,
+        future_seq_len: Optional[int] = None,
+        only_sample_tail: bool = False,
+        transform: Optional[Callable] = None,
+        use_libero_goal: bool = False,
+    ):
+        if future_conditional:
+            assert future_seq_len is not None, "must specify a future_seq_len"
+        self.dataset = dataset
+        self.window = window
+        self.action_window = action_window
+        self.vqbet_get_future_action_chunk = vqbet_get_future_action_chunk
+        self.future_conditional = future_conditional
+        self.min_future_sep = min_future_sep
+        self.future_seq_len = future_seq_len
+        self.only_sample_tail = only_sample_tail
+        self.transform = transform
+        self.slices = []
+        self.use_libero_goal = use_libero_goal
+        min_seq_length = np.inf
+        if vqbet_get_future_action_chunk:
+            min_window_required = window + action_window
+        else:
+            min_window_required = max(window, action_window)
+        for i in range(len(self.dataset)):  # type: ignore
+            T = self.dataset.get_seq_length(i)  # avoid reading actual seq (slow)
+            min_seq_length = min(T, min_seq_length)
+            if T - min_window_required < 0:
+                print(
+                    f"Ignored short sequence #{i}: len={T}, window={min_window_required}"
+                )
+            else:
+                self.slices += [
+                    (i, 0, end + 1) for end in range(window - 1)
+                ]  # slice indices follow convention [start, end)
+                self.slices += [
+                    (i, start, start + window)
+                    for start in range(T - min_window_required)
+                ]  # slice indices follow convention [start, end)
+
+        if min_seq_length < min_window_required:
+            print(
+                f"Ignored short sequences. To include all, set window <= {min_seq_length}."
+            )
+
+    def get_seq_length(self, idx: int) -> int:
+        if self.future_conditional:
+            return self.future_seq_len + self.window
+        else:
+            return self.window
+
+    def __len__(self):
+        return len(self.slices)
+
+    def __getitem__(self, idx):
+        i, start, end = self.slices[idx]
+        if end - start < self.window:
+            obs, act, *others = self.dataset[i]
+            obs = utils.inference.repeat_start_to_length(
+                obs[start:end], self.window, dim=0
+            )
+            act = utils.inference.repeat_start_to_length(
+                act[start : end - 1 + self.action_window],
+                self.window + self.action_window - 1,
+                dim=0,
+            )
+            values = [obs, act]
+        else:
+            values = [
+                self.dataset[i][0][start:end],
+                self.dataset[i][1][start : end - 1 + self.action_window],
+            ]
+
+        if self.use_libero_goal:
+            goals = self.dataset[i][2][start:end]
+            if end - start < self.window:
+                goals = utils.inference.repeat_start_to_length(
+                    goals, self.window, dim=0
+                )
+            values.append(goals)
+
+        # optionally apply transform
+        if self.transform is not None:
+            values = self.transform(values)
+        if len(values) == 2:  # placeholder goal
+            values.append(torch.ones([1, 1, 1]))
+        return tuple(values)

datasets/your_dataset.py

@@ -0,0 +1,22 @@
+import utils
+import torch
+import numpy as np
+from pathlib import Path
+from torch.utils.data import TensorDataset
+from datasets.core import TrajectoryDataset
+
+
+class YourTrajectoryDataset(TensorDataset, TrajectoryDataset):
+    def __init__(self, data_directory):
+        data_directory = Path(data_directory)
+
+    def get_seq_length(self, idx):
+        raise NotImplementedError
+
+    def get_frames(self, idx, frames):
+        raise NotImplementedError
+        # return obs / 255.0, actions, masks
+
+    def __getitem__(self, idx):
+        T = self.get_seq_length(idx)
+        return self.get_frames(idx, range(T))

envs/assets/block.urdf

@@ -0,0 +1,31 @@
+<?xml version="0.0" ?>
+<robot name="box.urdf">
+  <link name="baseLink">
+    <contact>
+      <lateral_friction value="1.0"/>
+      <rolling_friction value="0.0001"/>
+      <inertia_scaling value="3.0"/>
+    </contact>
+    <inertial>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+       <mass value=".01"/>
+       <inertia ixx="1" ixy="0" ixz="0" iyy="1" iyz="0" izz="1"/>
+    </inertial>
+    <visual>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+         <box size="0.04 0.04 0.04"/>
+      </geometry>
+      <material name="red">
+        <color rgba="1 0.3412 0.3490 1"/>
+      </material>
+    </visual>
+    <collision>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+             <box size="0.04 0.04 0.04"/>
+      </geometry>
+    </collision>
+  </link>
+</robot>
+

envs/assets/block2.urdf

@@ -0,0 +1,31 @@
+<?xml version="0.0" ?>
+<robot name="box2.urdf">
+  <link name="baseLink">
+    <contact>
+      <lateral_friction value="1.0"/>
+      <rolling_friction value="0.0001"/>
+      <inertia_scaling value="3.0"/>
+    </contact>
+    <inertial>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+       <mass value=".01"/>
+       <inertia ixx="1" ixy="0" ixz="0" iyy="1" iyz="0" izz="1"/>
+    </inertial>
+    <visual>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+         <box size="0.04 0.04 0.04"/>
+      </geometry>
+      <material name="red">
+        <color rgba="0.3412 1 0.3490 1"/>
+      </material>
+    </visual>
+    <collision>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+             <box size="0.04 0.04 0.04"/>
+      </geometry>
+    </collision>
+  </link>
+</robot>
+

envs/assets/blocks/blue_cube.urdf

@@ -0,0 +1,30 @@
+<?xml version="1.0" ?>
+<robot name="blue_cube.urdf">
+  <link name="baseLink">
+    <contact>
+      <lateral_friction value="0.5"/>
+      <rolling_friction value="0.0001"/>
+      <inertia_scaling value="1.0"/>
+    </contact>
+    <inertial>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+       <mass value=".01"/>
+       <inertia ixx="1" ixy="0" ixz="0" iyy="1" iyz="0" izz="1"/>
+    </inertial>
+    <visual>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <mesh filename="cube.obj" scale="1.0 1.0 1.0"/>
+      </geometry>
+       <material name="blue">
+        <color rgba="0.4 0.4 1.0 1"/>
+      </material>
+    </visual>
+    <collision>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <mesh filename="cube.obj" scale="1.0 1.0 1.0"/>
+      </geometry>
+    </collision>
+  </link>
+</robot>

envs/assets/blocks/cube.obj

@@ -0,0 +1,446 @@
+# Blender v2.92.0 OBJ File: ''
+# www.blender.org
+mtllib cube.mtl
+o square_blue_block_Cube.001
+v 0.000000 0.000000 -0.000000
+v 0.000000 0.038100 -0.000000
+v -0.014865 0.000000 0.018106
+v -0.016277 0.002032 0.019826
+v -0.015863 0.000595 0.019322
+v -0.019826 0.002032 0.016277
+v -0.018106 0.000000 0.014865
+v -0.019322 0.000595 0.015863
+v -0.018052 0.002032 0.019351
+v -0.016494 0.000000 0.017681
+v -0.017595 0.000595 0.018862
+v -0.019351 0.002032 0.018052
+v -0.017681 0.000000 0.016494
+v -0.018862 0.000595 0.017595
+v -0.018106 0.038100 0.014865
+v -0.019826 0.036068 0.016277
+v -0.019322 0.037505 0.015863
+v -0.016277 0.036068 0.019826
+v -0.014865 0.038100 0.018106
+v -0.015863 0.037505 0.019322
+v -0.017681 0.038100 0.016494
+v -0.019351 0.036068 0.018052
+v -0.018862 0.037505 0.017595
+v -0.016494 0.038100 0.017681
+v -0.018052 0.036068 0.019351
+v -0.017595 0.037505 0.018862
+v -0.018106 0.000000 -0.014865
+v -0.019826 0.002032 -0.016277
+v -0.019322 0.000595 -0.015863
+v -0.016277 0.002032 -0.019826
+v -0.014865 0.000000 -0.018106
+v -0.015863 0.000595 -0.019322
+v -0.017681 0.000000 -0.016494
+v -0.019351 0.002032 -0.018052
+v -0.018862 0.000595 -0.017595
+v -0.016494 0.000000 -0.017681
+v -0.018052 0.002032 -0.019351
+v -0.017595 0.000595 -0.018862
+v -0.014865 0.038100 -0.018106
+v -0.016277 0.036068 -0.019826
+v -0.015863 0.037505 -0.019322
+v -0.019826 0.036068 -0.016277
+v -0.018106 0.038100 -0.014865
+v -0.019322 0.037505 -0.015863
+v -0.018052 0.036068 -0.019351
+v -0.016494 0.038100 -0.017681
+v -0.017595 0.037505 -0.018862
+v -0.019351 0.036068 -0.018052
+v -0.017681 0.038100 -0.016494
+v -0.018862 0.037505 -0.017595
+v 0.018106 0.000000 0.014865
+v 0.019826 0.002032 0.016277
+v 0.019322 0.000595 0.015863
+v 0.016277 0.002032 0.019826
+v 0.014865 0.000000 0.018106
+v 0.015863 0.000595 0.019322
+v 0.017681 0.000000 0.016494
+v 0.019351 0.002032 0.018052
+v 0.018862 0.000595 0.017595
+v 0.016494 0.000000 0.017681
+v 0.018052 0.002032 0.019351
+v 0.017595 0.000595 0.018862
+v 0.014865 0.038100 0.018106
+v 0.016277 0.036068 0.019826
+v 0.015863 0.037505 0.019322
+v 0.019826 0.036068 0.016277
+v 0.018106 0.038100 0.014865
+v 0.019322 0.037505 0.015863
+v 0.018052 0.036068 0.019351
+v 0.016494 0.038100 0.017681
+v 0.017595 0.037505 0.018862
+v 0.019351 0.036068 0.018052
+v 0.017681 0.038100 0.016494
+v 0.018862 0.037505 0.017595
+v 0.014865 0.000000 -0.018106
+v 0.016277 0.002032 -0.019826
+v 0.015863 0.000595 -0.019322
+v 0.019826 0.002032 -0.016277
+v 0.018106 0.000000 -0.014865
+v 0.019322 0.000595 -0.015863
+v 0.016494 0.000000 -0.017681
+v 0.018052 0.002032 -0.019351
+v 0.017595 0.000595 -0.018862
+v 0.017681 0.000000 -0.016494
+v 0.019351 0.002032 -0.018052
+v 0.018862 0.000595 -0.017595
+v 0.018106 0.038100 -0.014865
+v 0.019826 0.036068 -0.016277
+v 0.019322 0.037505 -0.015863
+v 0.016277 0.036068 -0.019826
+v 0.014865 0.038100 -0.018106
+v 0.015863 0.037505 -0.019322
+v 0.019351 0.036068 -0.018052
+v 0.017681 0.038100 -0.016494
+v 0.018862 0.037505 -0.017595
+v 0.018052 0.036068 -0.019351
+v 0.016494 0.038100 -0.017681
+v 0.017595 0.037505 -0.018862
+vt 0.811987 0.285513
+vt 0.811986 0.268118
+vt 0.822561 0.276819
+vt 0.831255 0.287393
+vt 0.813860 0.287394
+vt 0.836073 0.319870
+vt 0.856069 0.319870
+vt 0.856069 0.320950
+vt 0.836073 0.320950
+vt 0.813867 0.266245
+vt 0.831262 0.266244
+vt 0.832202 0.266496
+vt 0.832884 0.267184
+vt 0.856069 0.297507
+vt 0.836074 0.297507
+vt 0.836074 0.278381
+vt 0.856070 0.278381
+vt 0.812238 0.267178
+vt 0.812926 0.266496
+vt 0.856070 0.275143
+vt 0.836074 0.275143
+vt 0.836074 0.256018
+vt 0.856070 0.256018
+vt 0.833135 0.268125
+vt 0.833136 0.285520
+vt 0.833136 0.260613
+vt 0.832884 0.261553
+vt 0.822561 0.251912
+vt 0.836074 0.342234
+vt 0.856070 0.342234
+vt 0.856070 0.343313
+vt 0.836074 0.343313
+vt 0.832196 0.287142
+vt 0.831262 0.241337
+vt 0.832202 0.241589
+vt 0.832884 0.286460
+vt 0.812238 0.286454
+vt 0.811986 0.243211
+vt 0.812238 0.242271
+vt 0.836073 0.300745
+vt 0.856069 0.300745
+vt 0.836073 0.298586
+vt 0.856069 0.298586
+vt 0.856069 0.299665
+vt 0.836073 0.299665
+vt 0.856070 0.276222
+vt 0.836074 0.276222
+vt 0.836074 0.277302
+vt 0.856070 0.277302
+vt 0.836074 0.253859
+vt 0.856070 0.253859
+vt 0.856070 0.254938
+vt 0.836074 0.254938
+vt 0.812920 0.287142
+vt 0.856069 0.322029
+vt 0.836073 0.322029
+vt 0.813860 0.262487
+vt 0.812920 0.262235
+vt 0.836073 0.323108
+vt 0.856069 0.323108
+vt 0.832884 0.242277
+vt 0.833135 0.243218
+vt 0.812238 0.261547
+vt 0.811987 0.260606
+vt 0.813868 0.241338
+vt 0.812926 0.241589
+vt 0.831255 0.262486
+vt 0.832196 0.262235
+vt 0.811196 0.267507
+vt 0.811506 0.266500
+vt 0.810258 0.267085
+vt 0.810726 0.265896
+vt 0.811523 0.262225
+vt 0.811189 0.261224
+vt 0.810732 0.262868
+vt 0.810258 0.261658
+vt 0.812248 0.265781
+vt 0.811605 0.264990
+vt 0.812242 0.262967
+vt 0.811638 0.263747
+vt 0.813249 0.265447
+vt 0.812815 0.264516
+vt 0.813249 0.263277
+vt 0.812827 0.264215
+vt 0.833926 0.261224
+vt 0.833616 0.262231
+vt 0.834864 0.261646
+vt 0.834396 0.262835
+vt 0.833599 0.266506
+vt 0.833933 0.267507
+vt 0.834390 0.265863
+vt 0.834864 0.267073
+vt 0.832874 0.262950
+vt 0.833517 0.263741
+vt 0.832880 0.265764
+vt 0.833484 0.264984
+vt 0.831873 0.263284
+vt 0.832307 0.264215
+vt 0.831873 0.265453
+vt 0.832295 0.264516
+vt 0.831873 0.240547
+vt 0.832880 0.240857
+vt 0.832295 0.239609
+vt 0.833484 0.240077
+vt 0.832874 0.287857
+vt 0.831873 0.288191
+vt 0.833517 0.288648
+vt 0.832307 0.289122
+vt 0.833599 0.241599
+vt 0.834390 0.240956
+vt 0.833616 0.287138
+vt 0.834396 0.287742
+vt 0.833933 0.242600
+vt 0.834864 0.242166
+vt 0.833926 0.286131
+vt 0.834864 0.286553
+vt 0.811196 0.242600
+vt 0.811506 0.241593
+vt 0.810258 0.242178
+vt 0.810726 0.240989
+vt 0.811523 0.287132
+vt 0.811189 0.286131
+vt 0.810732 0.287775
+vt 0.810258 0.286565
+vt 0.812248 0.240874
+vt 0.811605 0.240083
+vt 0.812242 0.287874
+vt 0.811638 0.288654
+vt 0.813249 0.240540
+vt 0.812815 0.239609
+vt 0.813249 0.288184
+vt 0.812827 0.289122
+vn 0.0000 -1.0000 0.0000
+vn 0.9739 0.1816 -0.1363
+vn 0.9739 -0.1816 -0.1363
+vn 0.8444 -0.2052 -0.4948
+vn 0.8444 0.2052 -0.4948
+vn 0.1363 -0.1816 0.9739
+vn 0.1363 0.1816 0.9739
+vn -0.1363 0.1816 0.9739
+vn -0.1363 -0.1816 0.9739
+vn -0.9739 -0.1816 0.1362
+vn -0.9739 0.1816 0.1362
+vn -0.9739 0.1816 -0.1362
+vn -0.9739 -0.1816 -0.1362
+vn -0.0000 1.0000 -0.0000
+vn -0.1363 0.1816 -0.9739
+vn -0.1363 -0.1816 -0.9739
+vn -0.4948 -0.2052 -0.8444
+vn -0.4948 0.2052 -0.8444
+vn 0.9739 0.1816 0.1363
+vn 0.9739 -0.1816 0.1363
+vn 0.4948 0.2052 0.8444
+vn 0.4948 -0.2052 0.8444
+vn 0.8444 -0.2052 0.4948
+vn 0.8444 0.2052 0.4948
+vn -0.8444 -0.2052 0.4948
+vn -0.8444 0.2052 0.4948
+vn -0.4948 0.2052 0.8444
+vn -0.4948 -0.2052 0.8444
+vn -0.8444 -0.2052 -0.4948
+vn -0.8444 0.2052 -0.4948
+vn 0.4948 -0.2052 -0.8444
+vn 0.4948 0.2052 -0.8444
+vn 0.1363 0.1816 -0.9739
+vn 0.1363 -0.1816 -0.9739
+vn -0.0965 -0.9775 0.1874
+vn -0.0187 -0.9731 0.2298
+vn -0.0935 -0.6743 0.7325
+vn -0.3529 -0.6994 0.6215
+vn -0.0187 0.9731 0.2298
+vn -0.0965 0.9775 0.1874
+vn -0.3529 0.6994 0.6215
+vn -0.0935 0.6743 0.7325
+vn -0.1874 -0.9775 0.0965
+vn -0.6215 -0.6994 0.3529
+vn -0.1874 0.9775 0.0965
+vn -0.6215 0.6994 0.3529
+vn -0.2298 -0.9731 0.0187
+vn -0.7325 -0.6743 0.0935
+vn -0.2298 0.9731 0.0187
+vn -0.7325 0.6743 0.0935
+vn -0.0965 0.9775 -0.1874
+vn -0.0187 0.9731 -0.2298
+vn -0.0935 0.6743 -0.7325
+vn -0.3529 0.6994 -0.6215
+vn -0.0187 -0.9731 -0.2298
+vn -0.0965 -0.9775 -0.1874
+vn -0.3529 -0.6994 -0.6215
+vn -0.0935 -0.6743 -0.7325
+vn -0.1874 0.9775 -0.0965
+vn -0.6215 0.6994 -0.3529
+vn -0.1874 -0.9775 -0.0965
+vn -0.6215 -0.6994 -0.3529
+vn -0.2298 0.9731 -0.0187
+vn -0.7325 0.6743 -0.0935
+vn -0.2298 -0.9731 -0.0187
+vn -0.7325 -0.6743 -0.0935
+vn 0.1874 0.9775 -0.0965
+vn 0.2298 0.9731 -0.0187
+vn 0.7325 0.6743 -0.0935
+vn 0.6215 0.6994 -0.3529
+vn 0.2298 -0.9731 -0.0187
+vn 0.1874 -0.9775 -0.0965
+vn 0.6215 -0.6994 -0.3529
+vn 0.7325 -0.6743 -0.0935
+vn 0.0965 0.9775 -0.1874
+vn 0.3529 0.6994 -0.6215
+vn 0.0965 -0.9775 -0.1874
+vn 0.3529 -0.6994 -0.6215
+vn 0.0187 0.9731 -0.2298
+vn 0.0935 0.6743 -0.7325
+vn 0.0187 -0.9731 -0.2298
+vn 0.0935 -0.6743 -0.7325
+vn 0.0965 0.9775 0.1874
+vn 0.0187 0.9731 0.2298
+vn 0.0935 0.6743 0.7325
+vn 0.3529 0.6994 0.6215
+vn 0.0187 -0.9731 0.2298
+vn 0.0965 -0.9775 0.1874
+vn 0.3529 -0.6994 0.6215
+vn 0.0935 -0.6743 0.7325
+vn 0.1874 0.9775 0.0965
+vn 0.6215 0.6994 0.3529
+vn 0.1874 -0.9775 0.0965
+vn 0.6215 -0.6994 0.3529
+vn 0.2298 0.9731 0.0187
+vn 0.7325 0.6743 0.0935
+vn 0.2298 -0.9731 0.0187
+vn 0.7325 -0.6743 0.0935
+usemtl toybox.001
+s 1
+f 55/1/1 3/2/1 1/3/1
+f 79/4/1 51/5/1 1/3/1
+f 88/6/2 78/7/3 85/8/4 93/9/5
+f 7/10/1 27/11/1 1/3/1
+f 33/12/1 36/13/1 1/3/1
+f 54/14/6 64/15/7 18/16/8 4/17/9
+f 10/18/1 13/19/1 1/3/1
+f 6/20/10 16/21/11 42/22/12 28/23/13
+f 31/24/1 75/25/1 1/3/1
+f 39/26/14 46/27/14 2/28/14
+f 40/29/15 30/30/16 37/31/17 45/32/18
+f 84/33/1 79/4/1 1/3/1
+f 87/34/14 94/35/14 2/28/14
+f 75/25/1 81/36/1 1/3/1
+f 60/37/1 55/1/1 1/3/1
+f 63/38/14 70/39/14 2/28/14
+f 78/7/3 88/6/2 66/40/19 52/41/20
+f 69/42/21 61/43/22 58/44/23 72/45/24
+f 13/19/1 7/10/1 1/3/1
+f 12/46/25 22/47/26 16/21/11 6/20/10
+f 81/36/1 84/33/1 1/3/1
+f 4/17/9 18/16/8 25/48/27 9/49/28
+f 45/50/18 37/51/17 34/52/29 48/53/30
+f 72/45/24 58/44/23 52/41/20 66/40/19
+f 57/54/1 60/37/1 1/3/1
+f 93/9/5 85/8/4 82/55/31 96/56/32
+f 9/49/28 25/48/27 22/47/26 12/46/25
+f 48/53/30 34/52/29 28/23/13 42/22/12
+f 36/13/1 31/24/1 1/3/1
+f 15/57/14 21/58/14 2/28/14
+f 30/30/16 40/29/15 90/59/33 76/60/34
+f 64/15/7 54/14/6 61/43/22 69/42/21
+f 97/61/14 91/62/14 2/28/14
+f 51/5/1 57/54/1 1/3/1
+f 24/63/14 19/64/14 2/28/14
+f 67/65/14 87/34/14 2/28/14
+f 73/66/14 67/65/14 2/28/14
+f 21/58/14 24/63/14 2/28/14
+f 43/67/14 15/57/14 2/28/14
+f 70/39/14 73/66/14 2/28/14
+f 19/64/14 63/38/14 2/28/14
+f 46/27/14 49/68/14 2/28/14
+f 49/68/14 43/67/14 2/28/14
+f 91/62/14 39/26/14 2/28/14
+f 96/56/32 82/55/31 76/60/34 90/59/33
+f 94/35/14 97/61/14 2/28/14
+f 27/11/1 33/12/1 1/3/1
+f 10/18/35 3/2/36 5/69/37 11/70/38
+f 11/70/38 5/69/37 4/71/9 9/72/28
+f 19/64/39 24/63/40 26/73/41 20/74/42
+f 20/74/42 26/73/41 25/75/27 18/76/8
+f 13/19/43 10/18/35 11/70/38 14/77/44
+f 14/77/44 11/70/38 9/72/28 12/78/25
+f 24/63/40 21/58/45 23/79/46 26/73/41
+f 26/73/41 23/79/46 22/80/26 25/75/27
+f 7/10/47 13/19/43 14/77/44 8/81/48
+f 8/81/48 14/77/44 12/78/25 6/82/10
+f 21/58/45 15/57/49 17/83/50 23/79/46
+f 23/79/46 17/83/50 16/84/11 22/80/26
+f 46/27/51 39/26/52 41/85/53 47/86/54
+f 47/86/54 41/85/53 40/87/15 45/88/18
+f 31/24/55 36/13/56 38/89/57 32/90/58
+f 32/90/58 38/89/57 37/91/17 30/92/16
+f 49/68/59 46/27/51 47/86/54 50/93/60
+f 50/93/60 47/86/54 45/88/18 48/94/30
+f 36/13/56 33/12/61 35/95/62 38/89/57
+f 38/89/57 35/95/62 34/96/29 37/91/17
+f 43/67/63 49/68/59 50/93/60 44/97/64
+f 44/97/64 50/93/60 48/94/30 42/98/12
+f 33/12/61 27/11/65 29/99/66 35/95/62
+f 35/95/62 29/99/66 28/100/13 34/96/29
+f 94/35/67 87/34/68 89/101/69 95/102/70
+f 95/102/70 89/101/69 88/103/2 93/104/5
+f 79/4/71 84/33/72 86/105/73 80/106/74
+f 80/106/74 86/105/73 85/107/4 78/108/3
+f 97/61/75 94/35/67 95/102/70 98/109/76
+f 98/109/76 95/102/70 93/104/5 96/110/32
+f 84/33/72 81/36/77 83/111/78 86/105/73
+f 86/105/73 83/111/78 82/112/31 85/107/4
+f 91/62/79 97/61/75 98/109/76 92/113/80
+f 92/113/80 98/109/76 96/110/32 90/114/33
+f 81/36/77 75/25/81 77/115/82 83/111/78
+f 83/111/78 77/115/82 76/116/34 82/112/31
+f 70/39/83 63/38/84 65/117/85 71/118/86
+f 71/118/86 65/117/85 64/119/7 69/120/21
+f 55/1/87 60/37/88 62/121/89 56/122/90
+f 56/122/90 62/121/89 61/123/22 54/124/6
+f 73/66/91 70/39/83 71/118/86 74/125/92
+f 74/125/92 71/118/86 69/120/21 72/126/24
+f 60/37/88 57/54/93 59/127/94 62/121/89
+f 62/121/89 59/127/94 58/128/23 61/123/22
+f 67/65/95 73/66/91 74/125/92 68/129/96
+f 68/129/96 74/125/92 72/126/24 66/130/19
+f 57/54/93 51/5/97 53/131/98 59/127/94
+f 59/127/94 53/131/98 52/132/20 58/128/23
+f 6/82/10 28/100/13 29/99/66 8/81/48
+f 8/81/48 29/99/66 27/11/65 7/10/47
+f 30/92/16 76/116/34 77/115/82 32/90/58
+f 32/90/58 77/115/82 75/25/81 31/24/55
+f 78/108/3 52/132/20 53/131/98 80/106/74
+f 80/106/74 53/131/98 51/5/97 79/4/71
+f 54/124/6 4/71/9 5/69/37 56/122/90
+f 56/122/90 5/69/37 3/2/36 55/1/87
+f 15/57/49 43/67/63 44/97/64 17/83/50
+f 17/83/50 44/97/64 42/98/12 16/84/11
+f 63/38/84 19/64/39 20/74/42 65/117/85
+f 65/117/85 20/74/42 18/76/8 64/119/7
+f 66/130/19 88/103/2 89/101/69 68/129/96
+f 68/129/96 89/101/69 87/34/68 67/65/95
+f 90/114/33 40/87/15 41/85/53 92/113/80
+f 92/113/80 41/85/53 39/26/52 91/62/79
+f 3/2/1 10/18/1 1/3/1

envs/assets/blocks/green_star.urdf

@@ -0,0 +1,30 @@
+<?xml version="1.0" ?>
+<robot name="green_star.urdf">
+  <link name="baseLink">
+    <contact>
+      <lateral_friction value="0.5"/>
+      <rolling_friction value="0.0001"/>
+      <inertia_scaling value="1.0"/>
+    </contact>
+    <inertial>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+       <mass value=".01"/>
+       <inertia ixx="1" ixy="0" ixz="0" iyy="1" iyz="0" izz="1"/>
+    </inertial>
+    <visual>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <mesh filename="star.obj" scale="1.0 1.0 1.0"/>
+      </geometry>
+       <material name="green">
+        <color rgba="0.4 1.0 0.4 1"/>
+      </material>
+    </visual>
+    <collision>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <mesh filename="star.obj" scale="1.0 1.0 1.0"/>
+      </geometry>
+    </collision>
+  </link>
+</robot>

envs/assets/blocks/moon.obj

@@ -0,0 +1,446 @@
+# Blender v2.92.0 OBJ File: ''
+# www.blender.org
+mtllib block.mtl
+o moon_red_block_Cylinder.012
+v -0.009181 0.000000 -0.006901
+v -0.012205 0.002032 -0.009549
+v -0.011319 0.000595 -0.008773
+v -0.021856 0.000000 -0.011560
+v -0.021140 0.002032 -0.012614
+v -0.021633 0.000595 -0.012305
+v -0.021140 0.036068 -0.012614
+v -0.021856 0.038100 -0.011560
+v -0.021633 0.037505 -0.012305
+v -0.014039 0.002032 0.009151
+v -0.012620 0.000000 0.008341
+v -0.013624 0.000595 0.008914
+v -0.012620 0.038100 0.008341
+v -0.014039 0.036068 0.009151
+v -0.013624 0.037505 0.008914
+v -0.024317 0.002032 -0.000986
+v -0.021966 0.000000 -0.000780
+v -0.023628 0.000595 -0.000925
+v -0.012205 0.036068 -0.009549
+v -0.009181 0.038100 -0.006901
+v -0.011319 0.037505 -0.008773
+v -0.021966 0.038100 -0.000780
+v -0.024317 0.036068 -0.000986
+v -0.023628 0.037505 -0.000925
+v -0.023768 0.000000 -0.007207
+v -0.026243 0.002032 -0.008076
+v -0.025518 0.000595 -0.007821
+v -0.023815 0.000000 -0.010019
+v -0.026267 0.002032 -0.011168
+v -0.025549 0.000595 -0.010832
+v -0.024498 0.002032 -0.012861
+v -0.023166 0.000000 -0.011548
+v -0.023825 0.000595 -0.012476
+v -0.026243 0.036068 -0.008076
+v -0.023768 0.038100 -0.007207
+v -0.025518 0.037505 -0.007821
+v -0.023166 0.038100 -0.011548
+v -0.024498 0.036068 -0.012861
+v -0.023825 0.037505 -0.012476
+v -0.026267 0.036068 -0.011168
+v -0.023815 0.038100 -0.010019
+v -0.025549 0.037505 -0.010832
+v -0.000000 0.038100 0.001138
+v -0.000000 0.000000 0.001138
+v -0.000000 0.000000 -0.006070
+v -0.000000 0.002032 -0.008427
+v -0.000000 0.000595 -0.007737
+v -0.000000 0.036068 -0.008427
+v -0.000000 0.038100 -0.006070
+v -0.000000 0.037505 -0.007737
+v 0.009181 0.000000 -0.006901
+v 0.012205 0.002032 -0.009549
+v 0.011319 0.000595 -0.008773
+v 0.021856 0.000000 -0.011560
+v 0.021140 0.002032 -0.012614
+v 0.021633 0.000595 -0.012305
+v 0.021140 0.036068 -0.012614
+v 0.021856 0.038100 -0.011560
+v 0.021633 0.037505 -0.012305
+v -0.000000 0.002032 0.012861
+v -0.000000 0.000000 0.011641
+v -0.000000 0.000595 0.012504
+v -0.000000 0.038100 0.011640
+v -0.000000 0.036068 0.012861
+v -0.000000 0.037505 0.012503
+v 0.014039 0.002032 0.009151
+v 0.012620 0.000000 0.008341
+v 0.013624 0.000595 0.008914
+v 0.012620 0.038100 0.008341
+v 0.014039 0.036068 0.009151
+v 0.013624 0.037505 0.008914
+v 0.024317 0.002032 -0.000986
+v 0.021966 0.000000 -0.000780
+v 0.023628 0.000595 -0.000925
+v 0.012205 0.036068 -0.009549
+v 0.009181 0.038100 -0.006901
+v 0.011319 0.037505 -0.008773
+v 0.021966 0.038100 -0.000780
+v 0.024317 0.036068 -0.000986
+v 0.023628 0.037505 -0.000925
+v 0.023768 0.000000 -0.007207
+v 0.026243 0.002032 -0.008076
+v 0.025518 0.000595 -0.007821
+v 0.023815 0.000000 -0.010019
+v 0.026267 0.002032 -0.011168
+v 0.025549 0.000595 -0.010832
+v 0.024498 0.002032 -0.012861
+v 0.023166 0.000000 -0.011548
+v 0.023825 0.000595 -0.012476
+v 0.026243 0.036068 -0.008076
+v 0.023768 0.038100 -0.007207
+v 0.025518 0.037505 -0.007821
+v 0.023166 0.038100 -0.011548
+v 0.024498 0.036068 -0.012861
+v 0.023825 0.037505 -0.012476
+v 0.026267 0.036068 -0.011168
+v 0.023815 0.038100 -0.010019
+v 0.025549 0.037505 -0.010832
+vt 0.710579 0.167405
+vt 0.710864 0.161788
+vt 0.714864 0.161788
+vt 0.686288 0.163021
+vt 0.706306 0.163021
+vt 0.706306 0.165002
+vt 0.686288 0.165002
+vt 0.708556 0.175784
+vt 0.708475 0.175088
+vt 0.721180 0.161788
+vt 0.719512 0.169043
+vt 0.714650 0.174731
+vt 0.711007 0.176048
+vt 0.720186 0.145021
+vt 0.718589 0.144931
+vt 0.714732 0.130671
+vt 0.709411 0.176139
+vt 0.721121 0.143971
+vt 0.721041 0.144667
+vt 0.686288 0.138411
+vt 0.706306 0.138411
+vt 0.706306 0.146951
+vt 0.686288 0.146951
+vt 0.706306 0.155441
+vt 0.686288 0.155441
+vt 0.706306 0.159762
+vt 0.686288 0.159762
+vt 0.714946 0.143614
+vt 0.686288 0.177766
+vt 0.686288 0.170557
+vt 0.706306 0.170557
+vt 0.706306 0.177766
+vt 0.686288 0.161581
+vt 0.706306 0.161581
+vt 0.718732 0.130671
+vt 0.719018 0.136287
+vt 0.710084 0.137925
+vt 0.708416 0.130671
+vt 0.709447 0.168778
+vt 0.709791 0.161788
+vt 0.708408 0.169256
+vt 0.708838 0.161788
+vt 0.707932 0.175020
+vt 0.707041 0.174762
+vt 0.722722 0.161788
+vt 0.720733 0.170191
+vt 0.719979 0.169645
+vt 0.721789 0.161788
+vt 0.708863 0.139073
+vt 0.706874 0.130671
+vt 0.707807 0.130671
+vt 0.709618 0.138527
+vt 0.715051 0.176717
+vt 0.714702 0.175778
+vt 0.719806 0.130670
+vt 0.720149 0.137661
+vt 0.720758 0.130670
+vt 0.721188 0.138138
+vt 0.721665 0.143903
+vt 0.722555 0.143645
+vt 0.714546 0.145599
+vt 0.714894 0.144661
+vt 0.721564 0.145099
+vt 0.720661 0.146046
+vt 0.722312 0.145803
+vt 0.721098 0.146965
+vt 0.708935 0.177163
+vt 0.708032 0.176217
+vt 0.708498 0.178082
+vt 0.707284 0.176921
+vt 0.718902 0.146070
+vt 0.719002 0.147136
+vt 0.710695 0.177187
+vt 0.710594 0.178254
+vt 0.710579 0.156171
+vt 0.686288 0.192510
+vt 0.686288 0.190530
+vt 0.706306 0.190530
+vt 0.706306 0.192510
+vt 0.708556 0.147792
+vt 0.708475 0.148488
+vt 0.719512 0.154534
+vt 0.714650 0.148845
+vt 0.711007 0.147528
+vt 0.720186 0.116320
+vt 0.718589 0.116410
+vt 0.709411 0.147437
+vt 0.721121 0.117370
+vt 0.721041 0.116674
+vt 0.686288 0.129870
+vt 0.706306 0.129870
+vt 0.686288 0.121380
+vt 0.706306 0.121380
+vt 0.686288 0.117059
+vt 0.706306 0.117059
+vt 0.714946 0.117727
+vt 0.706306 0.184974
+vt 0.686288 0.184974
+vt 0.706306 0.193950
+vt 0.686288 0.193950
+vt 0.719018 0.125054
+vt 0.710084 0.123416
+vt 0.706306 0.115240
+vt 0.686288 0.115240
+vt 0.709447 0.154798
+vt 0.708408 0.154321
+vt 0.707932 0.148556
+vt 0.707041 0.148814
+vt 0.719979 0.153931
+vt 0.720733 0.153385
+vt 0.708863 0.122268
+vt 0.709618 0.122814
+vt 0.714702 0.147798
+vt 0.715051 0.146860
+vt 0.720149 0.123680
+vt 0.721188 0.123203
+vt 0.721665 0.117438
+vt 0.722555 0.117696
+vt 0.714546 0.115742
+vt 0.714894 0.116680
+vt 0.720661 0.115295
+vt 0.721564 0.116241
+vt 0.721098 0.114376
+vt 0.722312 0.115538
+vt 0.708032 0.147359
+vt 0.708935 0.146413
+vt 0.707284 0.146655
+vt 0.708498 0.145494
+vt 0.718901 0.115271
+vt 0.719002 0.114205
+vt 0.710695 0.146389
+vt 0.710595 0.145323
+vn 0.0000 -1.0000 -0.0000
+vn -0.3609 -0.1941 -0.9122
+vn -0.3609 0.1941 -0.9122
+vn 0.2092 0.1359 -0.9684
+vn 0.2092 -0.1359 -0.9684
+vn 0.0000 1.0000 -0.0000
+vn -0.0000 -0.1176 0.9931
+vn -0.0000 0.1176 0.9931
+vn -0.5010 0.1369 0.8546
+vn -0.5010 -0.1369 0.8546
+vn -0.8572 0.1847 0.4806
+vn -0.8572 -0.1847 0.4806
+vn -0.9678 0.2183 0.1257
+vn -0.9678 -0.2183 0.1257
+vn 0.0000 -0.2255 -0.9742
+vn 0.2279 -0.1771 -0.9575
+vn 0.2279 0.1771 -0.9575
+vn 0.0000 0.2255 -0.9742
+vn -0.8945 -0.2361 -0.3797
+vn -0.8945 0.2361 -0.3797
+vn 0.0000 -0.9870 -0.1606
+vn 0.0247 -0.9852 -0.1694
+vn 0.1624 -0.6740 -0.7206
+vn 0.0000 -0.7551 -0.6556
+vn 0.0583 -0.9460 -0.3190
+vn 0.1517 -0.5680 -0.8089
+vn -0.4024 -0.5836 0.7053
+vn 0.0000 -0.5393 0.8421
+vn -0.1318 -0.9470 0.2929
+vn 0.0000 -0.9390 0.3440
+vn 0.0000 0.5393 0.8421
+vn -0.4024 0.5836 0.7053
+vn 0.0000 0.9390 0.3440
+vn -0.1318 0.9470 0.2929
+vn -0.6387 -0.6655 0.3863
+vn -0.2030 -0.9669 0.1548
+vn 0.0247 0.9852 -0.1694
+vn -0.0000 0.9870 -0.1606
+vn 0.0000 0.7551 -0.6556
+vn 0.1624 0.6740 -0.7206
+vn 0.0583 0.9460 -0.3190
+vn 0.1517 0.5680 -0.8089
+vn -0.6387 0.6655 0.3863
+vn -0.2030 0.9669 0.1548
+vn -0.1804 0.9821 -0.0543
+vn -0.0951 0.9531 -0.2874
+vn -0.2234 0.6362 -0.7384
+vn -0.6431 0.7203 -0.2600
+vn -0.0951 -0.9531 -0.2874
+vn -0.1804 -0.9821 -0.0543
+vn -0.6431 -0.7203 -0.2600
+vn -0.2234 -0.6362 -0.7384
+vn -0.1858 0.9820 0.0343
+vn -0.6801 0.7265 0.0987
+vn -0.1858 -0.9820 0.0343
+vn -0.6801 -0.7265 0.0987
+vn 0.3609 -0.1941 -0.9122
+vn -0.2092 -0.1359 -0.9684
+vn -0.2092 0.1359 -0.9684
+vn 0.3609 0.1941 -0.9122
+vn 0.5010 -0.1369 0.8546
+vn 0.5010 0.1369 0.8546
+vn 0.8572 -0.1847 0.4806
+vn 0.8572 0.1847 0.4806
+vn 0.9678 -0.2183 0.1257
+vn 0.9678 0.2183 0.1257
+vn -0.2279 0.1771 -0.9575
+vn -0.2279 -0.1771 -0.9575
+vn 0.8945 0.2361 -0.3797
+vn 0.8945 -0.2361 -0.3797
+vn -0.1624 -0.6740 -0.7206
+vn -0.0247 -0.9852 -0.1694
+vn -0.1517 -0.5680 -0.8089
+vn -0.0583 -0.9460 -0.3190
+vn 0.4024 -0.5836 0.7053
+vn 0.1318 -0.9470 0.2929
+vn 0.4024 0.5836 0.7053
+vn 0.1318 0.9470 0.2929
+vn 0.6387 -0.6655 0.3863
+vn 0.2030 -0.9669 0.1548
+vn -0.0247 0.9852 -0.1694
+vn -0.1624 0.6740 -0.7206
+vn -0.0583 0.9460 -0.3190
+vn -0.1517 0.5680 -0.8089
+vn 0.6387 0.6655 0.3863
+vn 0.2030 0.9669 0.1548
+vn 0.1804 0.9821 -0.0543
+vn 0.6431 0.7203 -0.2600
+vn 0.2234 0.6362 -0.7384
+vn 0.0951 0.9531 -0.2874
+vn 0.0951 -0.9531 -0.2874
+vn 0.2234 -0.6362 -0.7384
+vn 0.6431 -0.7203 -0.2600
+vn 0.1804 -0.9821 -0.0543
+vn 0.1858 0.9820 0.0343
+vn 0.6801 0.7265 0.0987
+vn 0.6801 -0.7265 0.0987
+vn 0.1858 -0.9820 0.0343
+usemtl toybox
+s 1
+f 1/1/1 45/2/1 44/3/1
+f 31/4/2 38/5/3 7/6/4 5/7/5
+f 32/8/1 4/9/1 44/3/1
+f 61/10/1 11/11/1 44/3/1
+f 11/11/1 17/12/1 44/3/1
+f 17/12/1 25/13/1 44/3/1
+f 41/14/6 35/15/6 43/16/6
+f 28/17/1 32/8/1 44/3/1
+f 8/18/6 37/19/6 43/16/6
+f 37/19/6 41/14/6 43/16/6
+f 25/13/1 28/17/1 44/3/1
+f 60/20/7 64/21/8 14/22/9 10/23/10
+f 10/23/10 14/22/9 23/24/11 16/25/12
+f 16/25/12 23/24/11 34/26/13 26/27/14
+f 35/15/6 22/28/6 43/16/6
+f 4/9/1 1/1/1 44/3/1
+f 46/29/15 2/30/16 19/31/17 48/32/18
+f 38/5/3 31/4/2 29/33/19 40/34/20
+f 49/35/6 20/36/6 43/16/6
+f 13/37/6 63/38/6 43/16/6
+f 2/30/16 5/7/5 7/6/4 19/31/17
+f 40/34/20 29/33/19 26/27/14 34/26/13
+f 20/36/6 8/18/6 43/16/6
+f 22/28/6 13/37/6 43/16/6
+f 45/2/21 1/1/22 3/39/23 47/40/24
+f 47/40/24 3/39/23 2/41/16 46/42/15
+f 1/1/22 4/9/25 6/43/26 3/39/23
+f 3/39/23 6/43/26 5/44/5 2/41/16
+f 60/45/7 10/46/10 12/47/27 62/48/28
+f 62/48/28 12/47/27 11/11/29 61/10/30
+f 14/49/9 64/50/8 65/51/31 15/52/32
+f 15/52/32 65/51/31 63/38/33 13/37/34
+f 10/46/10 16/53/12 18/54/35 12/47/27
+f 12/47/27 18/54/35 17/12/36 11/11/29
+f 20/36/37 49/35/38 50/55/39 21/56/40
+f 21/56/40 50/55/39 48/57/18 19/58/17
+f 8/18/41 20/36/37 21/56/40 9/59/42
+f 9/59/42 21/56/40 19/58/17 7/60/4
+f 23/61/11 14/49/9 15/52/32 24/62/43
+f 24/62/43 15/52/32 13/37/34 22/28/44
+f 41/14/45 37/19/46 39/63/47 42/64/48
+f 42/64/48 39/63/47 38/65/3 40/66/20
+f 32/8/49 28/17/50 30/67/51 33/68/52
+f 33/68/52 30/67/51 29/69/19 31/70/2
+f 35/15/53 41/14/45 42/64/48 36/71/54
+f 36/71/54 42/64/48 40/66/20 34/72/13
+f 28/17/50 25/13/55 27/73/56 30/67/51
+f 30/67/51 27/73/56 26/74/14 29/69/19
+f 25/13/55 17/12/36 18/54/35 27/73/56
+f 27/73/56 18/54/35 16/53/12 26/74/14
+f 22/28/44 35/15/53 36/71/54 24/62/43
+f 24/62/43 36/71/54 34/72/13 23/61/11
+f 4/9/25 32/8/49 33/68/52 6/43/26
+f 6/43/26 33/68/52 31/70/2 5/44/5
+f 37/19/46 8/18/41 9/59/42 39/63/47
+f 39/63/47 9/59/42 7/60/4 38/65/3
+f 51/75/1 44/3/1 45/2/1
+f 87/76/57 55/77/58 57/78/59 94/79/60
+f 88/80/1 44/3/1 54/81/1
+f 61/10/1 44/3/1 67/82/1
+f 67/82/1 44/3/1 73/83/1
+f 73/83/1 44/3/1 81/84/1
+f 97/85/6 43/16/6 91/86/6
+f 84/87/1 44/3/1 88/80/1
+f 58/88/6 43/16/6 93/89/6
+f 93/89/6 43/16/6 97/85/6
+f 81/84/1 44/3/1 84/87/1
+f 60/20/7 66/90/61 70/91/62 64/21/8
+f 66/90/61 72/92/63 79/93/64 70/91/62
+f 72/92/63 82/94/65 90/95/66 79/93/64
+f 91/86/6 43/16/6 78/96/6
+f 54/81/1 44/3/1 51/75/1
+f 46/29/15 48/32/18 75/97/67 52/98/68
+f 94/79/60 96/99/69 85/100/70 87/76/57
+f 49/35/6 43/16/6 76/101/6
+f 69/102/6 43/16/6 63/38/6
+f 52/98/68 75/97/67 57/78/59 55/77/58
+f 96/103/69 90/95/66 82/94/65 85/104/70
+f 76/101/6 43/16/6 58/88/6
+f 78/96/6 43/16/6 69/102/6
+f 45/2/21 47/40/24 53/105/71 51/75/72
+f 47/40/24 46/42/15 52/106/68 53/105/71
+f 51/75/72 53/105/71 56/107/73 54/81/74
+f 53/105/71 52/106/68 55/108/58 56/107/73
+f 60/45/7 62/48/28 68/109/75 66/110/61
+f 62/48/28 61/10/30 67/82/76 68/109/75
+f 70/111/62 71/112/77 65/51/31 64/50/8
+f 71/112/77 69/102/78 63/38/33 65/51/31
+f 66/110/61 68/109/75 74/113/79 72/114/63
+f 68/109/75 67/82/76 73/83/80 74/113/79
+f 76/101/81 77/115/82 50/55/39 49/35/38
+f 77/115/82 75/116/67 48/57/18 50/55/39
+f 58/88/83 59/117/84 77/115/82 76/101/81
+f 59/117/84 57/118/59 75/116/67 77/115/82
+f 79/119/64 80/120/85 71/112/77 70/111/62
+f 80/120/85 78/96/86 69/102/78 71/112/77
+f 97/85/87 98/121/88 95/122/89 93/89/90
+f 98/121/88 96/123/69 94/124/60 95/122/89
+f 88/80/91 89/125/92 86/126/93 84/87/94
+f 89/125/92 87/127/57 85/128/70 86/126/93
+f 91/86/95 92/129/96 98/121/88 97/85/87
+f 92/129/96 90/130/66 96/123/69 98/121/88
+f 84/87/94 86/126/93 83/131/97 81/84/98
+f 86/126/93 85/128/70 82/132/65 83/131/97
+f 81/84/98 83/131/97 74/113/79 73/83/80
+f 83/131/97 82/132/65 72/114/63 74/113/79
+f 78/96/86 80/120/85 92/129/96 91/86/95
+f 80/120/85 79/119/64 90/130/66 92/129/96
+f 54/81/74 56/107/73 89/125/92 88/80/91
+f 56/107/73 55/108/58 87/127/57 89/125/92
+f 93/89/90 95/122/89 59/117/84 58/88/83
+f 95/122/89 94/124/60 57/118/59 59/117/84

envs/assets/blocks/pentagon.obj

@@ -0,0 +1,419 @@
+# Blender v2.92.0 OBJ File: ''
+# www.blender.org
+mtllib pentagon.mtl
+o pentagon_yellow_block_Cube.003
+v -0.000000 0.000000 -0.001329
+v -0.000000 0.038100 -0.001329
+v -0.010789 0.000000 0.015873
+v -0.011940 0.002032 0.017708
+v -0.011603 0.000595 0.017170
+v -0.014201 0.002032 0.016092
+v -0.012833 0.000000 0.014413
+v -0.013800 0.000595 0.015600
+v -0.013310 0.002032 0.017235
+v -0.012029 0.000000 0.015448
+v -0.012935 0.000595 0.016711
+v -0.012833 0.038100 0.014413
+v -0.014201 0.036068 0.016092
+v -0.013800 0.037505 0.015600
+v -0.011940 0.036068 0.017708
+v -0.010789 0.038100 0.015873
+v -0.011603 0.037505 0.017170
+v -0.012029 0.038100 0.015448
+v -0.013310 0.036068 0.017235
+v -0.012935 0.037505 0.016711
+v -0.020074 0.000000 -0.006755
+v -0.022213 0.002032 -0.007333
+v -0.021586 0.000595 -0.007164
+v -0.021345 0.002032 -0.009898
+v -0.019288 0.000000 -0.009072
+v -0.020743 0.000595 -0.009656
+v -0.022188 0.002032 -0.008754
+v -0.020054 0.000000 -0.008040
+v -0.021563 0.000595 -0.008545
+v -0.019288 0.038100 -0.009072
+v -0.021345 0.036068 -0.009898
+v -0.020743 0.037505 -0.009656
+v -0.022213 0.036068 -0.007333
+v -0.020074 0.038100 -0.006755
+v -0.021586 0.037505 -0.007164
+v -0.022188 0.036068 -0.008754
+v -0.020054 0.038100 -0.008040
+v -0.021563 0.037505 -0.008545
+v 0.012833 0.000000 0.014413
+v 0.014201 0.002032 0.016092
+v 0.013800 0.000595 0.015600
+v 0.011940 0.002032 0.017708
+v 0.010789 0.000000 0.015873
+v 0.011603 0.000595 0.017170
+v 0.012029 0.000000 0.015448
+v 0.013310 0.002032 0.017235
+v 0.012935 0.000595 0.016711
+v 0.010789 0.038100 0.015873
+v 0.011940 0.036068 0.017708
+v 0.011603 0.037505 0.017170
+v 0.014201 0.036068 0.016092
+v 0.012833 0.038100 0.014413
+v 0.013800 0.037505 0.015600
+v 0.013310 0.036068 0.017235
+v 0.012029 0.038100 0.015448
+v 0.012935 0.037505 0.016711
+v 0.019288 0.000000 -0.009072
+v 0.021345 0.002032 -0.009898
+v 0.020743 0.000595 -0.009656
+v 0.022213 0.002032 -0.007333
+v 0.020074 0.000000 -0.006755
+v 0.021586 0.000595 -0.007164
+v 0.022188 0.002032 -0.008754
+v 0.020054 0.000000 -0.008040
+v 0.021563 0.000595 -0.008545
+v 0.020074 0.038100 -0.006755
+v 0.022213 0.036068 -0.007333
+v 0.021586 0.037505 -0.007164
+v 0.021345 0.036068 -0.009898
+v 0.019288 0.038100 -0.009072
+v 0.020743 0.037505 -0.009656
+v 0.022188 0.036068 -0.008754
+v 0.020054 0.038100 -0.008040
+v 0.021563 0.037505 -0.008545
+v 0.001283 0.038100 -0.021099
+v 0.001420 0.036068 -0.023214
+v 0.001380 0.037505 -0.022595
+v -0.001420 0.036068 -0.023214
+v -0.001283 0.038100 -0.021099
+v -0.001380 0.037505 -0.022595
+v 0.000000 0.036068 -0.023607
+v -0.000000 0.038100 -0.021456
+v -0.000000 0.037505 -0.022977
+v -0.001283 0.000000 -0.021099
+v -0.001420 0.002032 -0.023214
+v -0.001380 0.000595 -0.022595
+v 0.001420 0.002032 -0.023214
+v 0.001283 0.000000 -0.021099
+v 0.001380 0.000595 -0.022595
+v -0.000000 0.000000 -0.021456
+v 0.000000 0.002032 -0.023607
+v -0.000000 0.000595 -0.022977
+vt 0.718822 0.029303
+vt 0.730852 0.033346
+vt 0.721614 0.040891
+vt 0.709480 0.040139
+vt 0.717419 0.029729
+vt 0.717181 0.051685
+vt 0.709481 0.041565
+vt 0.686374 0.033371
+vt 0.706477 0.033371
+vt 0.706477 0.034211
+vt 0.686374 0.034211
+vt 0.731711 0.034535
+vt 0.731731 0.047644
+vt 0.732155 0.013538
+vt 0.732386 0.014251
+vt 0.720023 0.014212
+vt 0.724218 0.025374
+vt 0.723581 0.025796
+vt 0.731459 0.033813
+vt 0.686374 0.065488
+vt 0.706477 0.065488
+vt 0.706477 0.066344
+vt 0.686374 0.066344
+vt 0.730865 0.048779
+vt 0.718604 0.052165
+vt 0.717826 0.052123
+vt 0.710784 0.021757
+vt 0.710178 0.021290
+vt 0.706477 0.050528
+vt 0.686374 0.050528
+vt 0.686374 0.049672
+vt 0.706477 0.049672
+vt 0.709250 0.040852
+vt 0.718055 0.029306
+vt 0.686374 0.051384
+vt 0.706477 0.051384
+vt 0.686374 0.035050
+vt 0.706477 0.035050
+vt 0.686374 0.019217
+vt 0.706477 0.019217
+vt 0.731482 0.048351
+vt 0.709906 0.007458
+vt 0.710155 0.006751
+vt 0.686374 0.003322
+vt 0.706477 0.003322
+vt 0.706477 0.017476
+vt 0.686374 0.017476
+vt 0.706477 0.081821
+vt 0.686374 0.081821
+vt 0.686374 0.067199
+vt 0.706477 0.067199
+vt 0.723811 0.002980
+vt 0.724455 0.003418
+vt 0.706477 0.082661
+vt 0.686374 0.082661
+vt 0.686374 0.002483
+vt 0.706477 0.002483
+vt 0.732157 0.014964
+vt 0.723033 0.002938
+vt 0.722814 0.025800
+vt 0.686374 0.018347
+vt 0.706477 0.018347
+vt 0.709926 0.020568
+vt 0.710771 0.006324
+vt 0.733509 0.048629
+vt 0.733072 0.049555
+vt 0.732273 0.048949
+vt 0.732572 0.048190
+vt 0.732296 0.050225
+vt 0.731620 0.049441
+vt 0.734168 0.013254
+vt 0.734379 0.014256
+vt 0.733377 0.014254
+vt 0.733156 0.013468
+vt 0.734166 0.015257
+vt 0.733156 0.015039
+vt 0.707468 0.041849
+vt 0.707257 0.040847
+vt 0.708259 0.040849
+vt 0.708480 0.041634
+vt 0.707471 0.039846
+vt 0.708480 0.040063
+vt 0.708128 0.006474
+vt 0.708564 0.005547
+vt 0.709364 0.006153
+vt 0.709064 0.006913
+vt 0.709341 0.004877
+vt 0.710016 0.005662
+vt 0.731599 0.032714
+vt 0.732251 0.033244
+vt 0.732301 0.031933
+vt 0.733073 0.032652
+vt 0.723869 0.026728
+vt 0.723033 0.026764
+vt 0.724166 0.027687
+vt 0.723129 0.027790
+vt 0.732546 0.034029
+vt 0.733509 0.033612
+vt 0.724576 0.026281
+vt 0.725089 0.027190
+vt 0.710038 0.022388
+vt 0.709386 0.021859
+vt 0.709336 0.023170
+vt 0.708563 0.022451
+vt 0.717767 0.028375
+vt 0.718604 0.028339
+vt 0.717470 0.027416
+vt 0.718508 0.027313
+vt 0.709090 0.021074
+vt 0.708128 0.021490
+vt 0.717060 0.028822
+vt 0.716547 0.027913
+vt 0.723265 0.002005
+vt 0.724118 0.002071
+vt 0.723385 0.000970
+vt 0.724440 0.001117
+vt 0.717519 0.053032
+vt 0.716801 0.052568
+vt 0.717197 0.053986
+vt 0.716269 0.053464
+vt 0.724836 0.002535
+vt 0.725367 0.001639
+vt 0.718372 0.053098
+vt 0.718251 0.054133
+vn 0.0000 -1.0000 0.0000
+vn -0.6839 0.1913 -0.7040
+vn -0.6839 -0.1913 -0.7040
+vn -0.9271 -0.2054 -0.3134
+vn -0.9271 0.2054 -0.3134
+vn 0.0000 1.0000 -0.0000
+vn 0.1696 0.1897 0.9671
+vn 0.1696 -0.1897 0.9671
+vn 0.5696 -0.2019 0.7968
+vn 0.5696 0.2019 0.7968
+vn -0.5696 -0.2019 0.7968
+vn -0.5696 0.2019 0.7968
+vn -0.8602 0.1898 0.4734
+vn -0.8602 -0.1898 0.4734
+vn -0.1696 0.1897 0.9671
+vn -0.1696 -0.1897 0.9671
+vn -0.9708 0.1916 0.1446
+vn -0.9708 -0.1916 0.1446
+vn -0.4034 0.1878 -0.8955
+vn -0.4034 -0.1878 -0.8955
+vn 0.6839 0.1913 -0.7040
+vn 0.6839 -0.1913 -0.7040
+vn 0.4034 -0.1878 -0.8955
+vn 0.4034 0.1878 -0.8955
+vn 0.9708 -0.1916 0.1446
+vn 0.9708 0.1916 0.1446
+vn 0.8602 0.1898 0.4734
+vn 0.8602 -0.1898 0.4734
+vn 0.9271 -0.2054 -0.3134
+vn 0.9271 0.2054 -0.3134
+vn 0.0000 0.1985 -0.9801
+vn 0.0000 -0.1985 -0.9801
+vn -0.6793 -0.6969 -0.2300
+vn -0.7180 -0.6862 0.1166
+vn -0.2003 -0.9774 -0.0681
+vn -0.2163 -0.9753 0.0454
+vn -0.5001 -0.6857 -0.5289
+vn -0.1444 -0.9752 -0.1679
+vn -0.6793 0.6969 -0.2300
+vn -0.5001 0.6857 -0.5289
+vn -0.2003 0.9774 -0.0681
+vn -0.1444 0.9752 -0.1679
+vn -0.7180 0.6862 0.1166
+vn -0.2163 0.9753 0.0454
+vn 0.6793 -0.6969 -0.2300
+vn 0.5001 -0.6857 -0.5289
+vn 0.2003 -0.9774 -0.0681
+vn 0.1444 -0.9752 -0.1679
+vn 0.7180 -0.6862 0.1166
+vn 0.2163 -0.9753 0.0454
+vn 0.6793 0.6969 -0.2300
+vn 0.7180 0.6862 0.1166
+vn 0.2003 0.9774 -0.0681
+vn 0.2163 0.9753 0.0454
+vn 0.5001 0.6857 -0.5289
+vn 0.1444 0.9752 -0.1679
+vn -0.1241 -0.9769 0.1739
+vn -0.0267 -0.9752 0.2199
+vn -0.1178 -0.6841 0.7198
+vn -0.4189 -0.6935 0.5862
+vn -0.0267 0.9752 0.2199
+vn -0.1242 0.9769 0.1739
+vn -0.4189 0.6935 0.5862
+vn -0.1178 0.6841 0.7198
+vn -0.1992 -0.9752 0.0965
+vn -0.6427 -0.6843 0.3444
+vn -0.1992 0.9752 0.0965
+vn -0.6427 0.6843 0.3444
+vn 0.1242 0.9769 0.1739
+vn 0.0267 0.9752 0.2199
+vn 0.1178 0.6841 0.7198
+vn 0.4189 0.6935 0.5862
+vn 0.0267 -0.9752 0.2199
+vn 0.1241 -0.9769 0.1739
+vn 0.4189 -0.6935 0.5862
+vn 0.1178 -0.6841 0.7198
+vn 0.1992 0.9752 0.0965
+vn 0.6427 0.6843 0.3444
+vn 0.1992 -0.9752 0.0965
+vn 0.6427 -0.6843 0.3444
+vn -0.0000 0.9764 -0.2159
+vn 0.1005 0.9750 -0.1981
+vn 0.3067 0.6821 -0.6638
+vn -0.0000 0.6900 -0.7238
+vn 0.1005 -0.9750 -0.1981
+vn 0.0000 -0.9764 -0.2159
+vn 0.0000 -0.6900 -0.7238
+vn 0.3067 -0.6821 -0.6638
+vn -0.1005 0.9750 -0.1981
+vn -0.3067 0.6821 -0.6638
+vn -0.1005 -0.9750 -0.1981
+vn -0.3067 -0.6821 -0.6638
+usemtl toybox
+s 1
+f 43/1/1 3/2/1 1/3/1
+f 61/4/1 39/5/1 1/3/1
+f 88/6/1 57/7/1 1/3/1
+f 31/8/2 24/9/3 27/10/4 36/11/5
+f 7/12/1 21/13/1 1/3/1
+f 30/14/6 37/15/6 2/16/6
+f 12/17/6 18/18/6 2/16/6
+f 10/19/1 7/12/1 1/3/1
+f 49/20/7 42/21/8 46/22/9 54/23/10
+f 25/24/1 84/25/1 1/3/1
+f 90/26/1 88/6/1 1/3/1
+f 3/2/1 10/19/1 1/3/1
+f 48/27/6 55/28/6 2/16/6
+f 9/29/11 19/30/12 13/31/13 6/32/14
+f 64/33/1 61/4/1 1/3/1
+f 39/5/1 45/34/1 1/3/1
+f 42/21/8 49/20/7 15/35/15 4/36/16
+f 6/32/14 13/31/13 33/37/17 22/38/18
+f 24/9/3 31/8/2 78/39/19 85/40/20
+f 21/13/1 28/41/1 1/3/1
+f 66/42/6 73/43/6 2/16/6
+f 69/44/21 58/45/22 87/46/23 76/47/24
+f 45/34/1 43/1/1 1/3/1
+f 60/48/25 67/49/26 51/50/27 40/51/28
+f 82/52/6 79/53/6 2/16/6
+f 84/25/1 90/26/1 1/3/1
+f 67/49/26 60/48/25 63/54/29 72/55/30
+f 28/41/1 25/24/1 1/3/1
+f 72/56/30 63/57/29 58/45/22 69/44/21
+f 37/15/6 34/58/6 2/16/6
+f 79/53/6 30/14/6 2/16/6
+f 75/59/6 82/52/6 2/16/6
+f 18/18/6 16/60/6 2/16/6
+f 54/23/10 46/22/9 40/51/28 51/50/27
+f 81/61/31 91/62/32 85/40/20 78/39/19
+f 16/60/6 48/27/6 2/16/6
+f 52/63/6 66/42/6 2/16/6
+f 36/11/5 27/10/4 22/38/18 33/37/17
+f 73/43/6 70/64/6 2/16/6
+f 57/7/1 64/33/1 1/3/1
+f 70/64/6 75/59/6 2/16/6
+f 34/58/6 12/17/6 2/16/6
+f 76/47/24 87/46/23 91/62/32 81/61/31
+f 55/28/6 52/63/6 2/16/6
+f 22/65/18 27/66/4 29/67/33 23/68/34
+f 23/68/34 29/67/33 28/41/35 21/13/36
+f 27/66/4 24/69/3 26/70/37 29/67/33
+f 29/67/33 26/70/37 25/24/38 28/41/35
+f 31/71/2 36/72/5 38/73/39 32/74/40
+f 32/74/40 38/73/39 37/15/41 30/14/42
+f 36/72/5 33/75/17 35/76/43 38/73/39
+f 38/73/39 35/76/43 34/58/44 37/15/41
+f 58/77/22 63/78/29 65/79/45 59/80/46
+f 59/80/46 65/79/45 64/33/47 57/7/48
+f 63/78/29 60/81/25 62/82/49 65/79/45
+f 65/79/45 62/82/49 61/4/50 64/33/47
+f 67/83/26 72/84/30 74/85/51 68/86/52
+f 68/86/52 74/85/51 73/43/53 66/42/54
+f 72/84/30 69/87/21 71/88/55 74/85/51
+f 74/85/51 71/88/55 70/64/56 73/43/53
+f 10/19/57 3/2/58 5/89/59 11/90/60
+f 11/90/60 5/89/59 4/91/16 9/92/11
+f 16/60/61 18/18/62 20/93/63 17/94/64
+f 17/94/64 20/93/63 19/95/12 15/96/15
+f 7/12/65 10/19/57 11/90/60 8/97/66
+f 8/97/66 11/90/60 9/92/11 6/98/14
+f 18/18/62 12/17/67 14/99/68 20/93/63
+f 20/93/63 14/99/68 13/100/13 19/95/12
+f 55/28/69 48/27/70 50/101/71 56/102/72
+f 56/102/72 50/101/71 49/103/7 54/104/10
+f 43/1/73 45/34/74 47/105/75 44/106/76
+f 44/106/76 47/105/75 46/107/9 42/108/8
+f 52/63/77 55/28/69 56/102/72 53/109/78
+f 53/109/78 56/102/72 54/104/10 51/110/27
+f 45/34/74 39/5/79 41/111/80 47/105/75
+f 47/105/75 41/111/80 40/112/28 46/107/9
+f 82/52/81 75/59/82 77/113/83 83/114/84
+f 83/114/84 77/113/83 76/115/24 81/116/31
+f 88/6/85 90/26/86 92/117/87 89/118/88
+f 89/118/88 92/117/87 91/119/32 87/120/23
+f 79/53/89 82/52/81 83/114/84 80/121/90
+f 80/121/90 83/114/84 81/116/31 78/122/19
+f 90/26/86 84/25/91 86/123/92 92/117/87
+f 92/117/87 86/123/92 85/124/20 91/119/32
+f 6/98/14 22/65/18 23/68/34 8/97/66
+f 8/97/66 23/68/34 21/13/36 7/12/65
+f 60/81/25 40/112/28 41/111/80 62/82/49
+f 62/82/49 41/111/80 39/5/79 61/4/50
+f 42/108/8 4/91/16 5/89/59 44/106/76
+f 44/106/76 5/89/59 3/2/58 43/1/73
+f 12/17/67 34/58/44 35/76/43 14/99/68
+f 14/99/68 35/76/43 33/75/17 13/100/13
+f 48/27/70 16/60/61 17/94/64 50/101/71
+f 50/101/71 17/94/64 15/96/15 49/103/7
+f 51/110/27 67/83/26 68/86/52 53/109/78
+f 53/109/78 68/86/52 66/42/54 52/63/77
+f 84/25/91 25/24/38 26/70/37 86/123/92
+f 86/123/92 26/70/37 24/69/3 85/124/20
+f 30/14/42 79/53/89 80/121/90 32/74/40
+f 32/74/40 80/121/90 78/122/19 31/71/2
+f 87/120/23 58/77/22 59/80/46 89/118/88
+f 89/118/88 59/80/46 57/7/48 88/6/85
+f 69/87/21 76/115/24 77/113/83 71/88/55
+f 71/88/55 77/113/83 75/59/82 70/64/56
+f 4/36/16 15/35/15 19/30/12 9/29/11

envs/assets/blocks/red_moon.urdf

@@ -0,0 +1,30 @@
+<?xml version="1.0" ?>
+<robot name="red_moon.urdf">
+  <link name="baseLink">
+    <contact>
+      <lateral_friction value="0.5"/>
+      <rolling_friction value="0.0001"/>
+      <inertia_scaling value="1.0"/>
+    </contact>
+    <inertial>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+       <mass value=".01"/>
+       <inertia ixx="1" ixy="0" ixz="0" iyy="1" iyz="0" izz="1"/>
+    </inertial>
+    <visual>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <mesh filename="moon.obj" scale="1.0 1.0 1.0"/>
+      </geometry>
+       <material name="red">
+        <color rgba="1 0.4 0.4 1"/>
+      </material>
+    </visual>
+    <collision>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <mesh filename="moon.obj" scale="1.0 1.0 1.0"/>
+      </geometry>
+    </collision>
+  </link>
+</robot>

envs/assets/blocks/star.obj

@@ -0,0 +1,689 @@
+# Blender v2.92.0 OBJ File: ''
+# www.blender.org
+mtllib star.mtl
+o star_green_block_star0_block
+v -0.000030 0.000000 0.001549
+v -0.000030 0.038100 0.001549
+v 0.006429 0.000000 -0.009092
+v 0.007380 0.002032 -0.010659
+v 0.007101 0.000595 -0.010200
+v 0.009909 0.002032 -0.008940
+v 0.008636 0.000000 -0.007597
+v 0.009536 0.000595 -0.008547
+v 0.008636 0.038100 -0.007597
+v 0.009909 0.036068 -0.008940
+v 0.009536 0.037505 -0.008547
+v 0.007380 0.036068 -0.010659
+v 0.006429 0.038100 -0.009092
+v 0.007101 0.037505 -0.010200
+v -0.006436 0.038100 -0.009091
+v -0.007380 0.036068 -0.010659
+v -0.007103 0.037505 -0.010200
+v -0.009909 0.036068 -0.008940
+v -0.008642 0.038100 -0.007595
+v -0.009538 0.037505 -0.008546
+v -0.008642 0.000000 -0.007595
+v -0.009909 0.002032 -0.008940
+v -0.009538 0.000595 -0.008546
+v -0.007380 0.002032 -0.010659
+v -0.006436 0.000000 -0.009091
+v -0.007103 0.000595 -0.010200
+v -0.011672 0.000000 0.006854
+v -0.013354 0.002032 0.007620
+v -0.012861 0.000595 0.007396
+v -0.014336 0.002032 0.004714
+v -0.012526 0.000000 0.004313
+v -0.013806 0.000595 0.004596
+v -0.012526 0.038100 0.004313
+v -0.014336 0.036068 0.004714
+v -0.013806 0.037505 0.004596
+v -0.013354 0.036068 0.007620
+v -0.011672 0.038100 0.006854
+v -0.012861 0.037505 0.007396
+v 0.012521 0.000000 0.004314
+v 0.014336 0.002032 0.004714
+v 0.013805 0.000595 0.004597
+v 0.013354 0.002032 0.007620
+v 0.011668 0.000000 0.006855
+v 0.012860 0.000595 0.007396
+v 0.011668 0.038100 0.006855
+v 0.013354 0.036068 0.007620
+v 0.012860 0.037505 0.007396
+v 0.014336 0.036068 0.004714
+v 0.012521 0.038100 0.004314
+v 0.013805 0.037505 0.004597
+v -0.010002 0.000000 0.018088
+v -0.011405 0.002032 0.020414
+v -0.010994 0.000595 0.019733
+v -0.013985 0.002032 0.018583
+v -0.012256 0.000000 0.016473
+v -0.013478 0.000595 0.017965
+v -0.013300 0.002032 0.020351
+v -0.011669 0.000000 0.018040
+v -0.012822 0.000595 0.019674
+v -0.012256 0.038100 0.016473
+v -0.013985 0.036068 0.018583
+v -0.013478 0.037505 0.017965
+v -0.011405 0.036068 0.020414
+v -0.010002 0.038100 0.018088
+v -0.010994 0.037505 0.019733
+v -0.011669 0.038100 0.018040
+v -0.013300 0.036068 0.020351
+v -0.012822 0.037505 0.019674
+v -0.018964 0.000000 -0.003158
+v -0.021672 0.002032 -0.003831
+v -0.020879 0.000595 -0.003634
+v -0.020696 0.002032 -0.006698
+v -0.018097 0.000000 -0.005661
+v -0.019935 0.000595 -0.006394
+v -0.019426 0.000000 -0.004715
+v -0.022190 0.002032 -0.005607
+v -0.021380 0.000595 -0.005346
+v -0.018097 0.038100 -0.005661
+v -0.020696 0.036068 -0.006698
+v -0.019935 0.037505 -0.006394
+v -0.021672 0.036068 -0.003831
+v -0.018964 0.038100 -0.003158
+v -0.020879 0.037505 -0.003634
+v -0.022190 0.036068 -0.005607
+v -0.019426 0.038100 -0.004715
+v -0.021380 0.037505 -0.005346
+v 0.012251 0.000000 0.016476
+v 0.013985 0.002032 0.018583
+v 0.013477 0.000595 0.017966
+v 0.011405 0.002032 0.020414
+v 0.009996 0.000000 0.018089
+v 0.010992 0.000595 0.019733
+v 0.011663 0.000000 0.018043
+v 0.013300 0.002032 0.020351
+v 0.012821 0.000595 0.019675
+v 0.009996 0.038100 0.018089
+v 0.011405 0.036068 0.020414
+v 0.010992 0.037505 0.019733
+v 0.013985 0.036068 0.018583
+v 0.012251 0.038100 0.016476
+v 0.013477 0.037505 0.017966
+v 0.013300 0.036068 0.020351
+v 0.011663 0.038100 0.018043
+v 0.012821 0.037505 0.019675
+v 0.018094 0.000000 -0.005663
+v 0.020696 0.002032 -0.006698
+v 0.019934 0.000595 -0.006395
+v 0.021672 0.002032 -0.003831
+v 0.018961 0.000000 -0.003159
+v 0.020878 0.000595 -0.003635
+v 0.019423 0.000000 -0.004716
+v 0.022190 0.002032 -0.005607
+v 0.021380 0.000595 -0.005346
+v 0.018961 0.038100 -0.003159
+v 0.021672 0.036068 -0.003831
+v 0.020878 0.037505 -0.003635
+v 0.020696 0.036068 -0.006698
+v 0.018094 0.038100 -0.005663
+v 0.019934 0.037505 -0.006395
+v 0.022190 0.036068 -0.005607
+v 0.019423 0.038100 -0.004716
+v 0.021380 0.037505 -0.005346
+v 0.001426 0.038100 -0.016730
+v 0.001638 0.036068 -0.019383
+v 0.001576 0.037505 -0.018606
+v -0.001638 0.036068 -0.019383
+v -0.001434 0.038100 -0.016729
+v -0.001578 0.037505 -0.018606
+v 0.000000 0.036068 -0.020414
+v -0.000004 0.038100 -0.017644
+v -0.000001 0.037505 -0.019603
+v -0.001434 0.000000 -0.016729
+v -0.001638 0.002032 -0.019383
+v -0.001578 0.000595 -0.018606
+v 0.001638 0.002032 -0.019383
+v 0.001426 0.000000 -0.016730
+v 0.001576 0.000595 -0.018606
+v -0.000004 0.000000 -0.017644
+v 0.000000 0.002032 -0.020414
+v -0.000001 0.000595 -0.019603
+v 0.001733 0.000000 0.015081
+v 0.001984 0.002032 0.017005
+v 0.001910 0.000595 0.016441
+v -0.001984 0.002032 0.017005
+v -0.001741 0.000000 0.015080
+v -0.001912 0.000595 0.016441
+v -0.001741 0.038100 0.015080
+v -0.001984 0.036068 0.017005
+v -0.001912 0.037505 0.016441
+v 0.001984 0.036068 0.017005
+v 0.001733 0.038100 0.015081
+v 0.001910 0.037505 0.016441
+vt 0.884326 0.143054
+vt 0.881637 0.147199
+vt 0.874485 0.145636
+vt 0.870501 0.139220
+vt 0.873534 0.134807
+vt 0.837942 0.112414
+vt 0.837942 0.131533
+vt 0.832314 0.131533
+vt 0.832314 0.112414
+vt 0.883572 0.123603
+vt 0.883611 0.122735
+vt 0.892897 0.129050
+vt 0.881240 0.148682
+vt 0.881539 0.153913
+vt 0.898370 0.134009
+vt 0.897030 0.139070
+vt 0.873770 0.153188
+vt 0.868788 0.154800
+vt 0.903052 0.129620
+vt 0.899208 0.132723
+vt 0.862147 0.112414
+vt 0.862147 0.131533
+vt 0.855821 0.131533
+vt 0.855821 0.112414
+vt 0.830184 0.112414
+vt 0.830184 0.131533
+vt 0.824015 0.131533
+vt 0.824015 0.112414
+vt 0.874942 0.134807
+vt 0.877932 0.139104
+vt 0.880430 0.154782
+vt 0.875317 0.153178
+vt 0.866663 0.147297
+vt 0.863946 0.143378
+vt 0.884646 0.135960
+vt 0.883837 0.135640
+vt 0.874236 0.134337
+vt 0.813888 0.131533
+vt 0.813888 0.112414
+vt 0.814927 0.112414
+vt 0.814927 0.131533
+vt 0.881273 0.154719
+vt 0.863186 0.112414
+vt 0.863186 0.131533
+vt 0.867919 0.154747
+vt 0.867585 0.153944
+vt 0.845799 0.112414
+vt 0.845799 0.131533
+vt 0.840171 0.131533
+vt 0.840171 0.112414
+vt 0.903484 0.128858
+vt 0.883832 0.141627
+vt 0.884500 0.142196
+vt 0.864368 0.141965
+vt 0.869278 0.140166
+vt 0.854097 0.112414
+vt 0.854097 0.131533
+vt 0.847929 0.131533
+vt 0.847929 0.112414
+vt 0.896527 0.139753
+vt 0.867312 0.149178
+vt 0.822292 0.112414
+vt 0.822292 0.131533
+vt 0.872132 0.131533
+vt 0.872132 0.112414
+vt 0.877999 0.112414
+vt 0.877999 0.131533
+vt 0.864225 0.131533
+vt 0.864225 0.112414
+vt 0.870414 0.112414
+vt 0.870414 0.131533
+vt 0.815966 0.112414
+vt 0.815966 0.131533
+vt 0.891348 0.136441
+vt 0.889875 0.135971
+vt 0.846864 0.131533
+vt 0.846864 0.112414
+vt 0.831249 0.131533
+vt 0.831249 0.112414
+vt 0.879159 0.140026
+vt 0.896695 0.118860
+vt 0.898202 0.123877
+vt 0.895355 0.118422
+vt 0.896169 0.118193
+vt 0.863717 0.142542
+vt 0.889642 0.122199
+vt 0.891100 0.121679
+vt 0.899083 0.125135
+vt 0.805981 0.112414
+vt 0.805981 0.131533
+vt 0.800114 0.131533
+vt 0.800114 0.112414
+vt 0.807699 0.112414
+vt 0.807699 0.131533
+vt 0.884410 0.122387
+vt 0.895707 0.139551
+vt 0.879086 0.112414
+vt 0.879086 0.131533
+vt 0.883766 0.134776
+vt 0.884979 0.130164
+vt 0.903027 0.128111
+vt 0.799027 0.131533
+vt 0.799027 0.112414
+vt 0.884945 0.128175
+vt 0.882437 0.147351
+vt 0.881976 0.149022
+vt 0.883261 0.147612
+vt 0.882837 0.149234
+vt 0.898778 0.123306
+vt 0.899783 0.124719
+vt 0.899515 0.122813
+vt 0.900468 0.124192
+vt 0.899922 0.133113
+vt 0.898965 0.134558
+vt 0.900624 0.133616
+vt 0.899718 0.135026
+vt 0.878303 0.138385
+vt 0.879696 0.139416
+vt 0.878850 0.137688
+vt 0.880184 0.138702
+vt 0.889274 0.121462
+vt 0.890920 0.120885
+vt 0.889004 0.120636
+vt 0.890592 0.120060
+vt 0.875420 0.153985
+vt 0.873676 0.154006
+vt 0.875394 0.154873
+vt 0.873704 0.154874
+vt 0.891195 0.137238
+vt 0.889532 0.136718
+vt 0.890894 0.138073
+vt 0.889289 0.137551
+vt 0.868700 0.139582
+vt 0.870085 0.138524
+vt 0.868187 0.138882
+vt 0.869517 0.137842
+vt 0.862942 0.143128
+vt 0.862743 0.142252
+vt 0.862000 0.143160
+vt 0.861880 0.141993
+vt 0.883020 0.136246
+vt 0.882908 0.135355
+vt 0.882298 0.136785
+vt 0.882012 0.135647
+vt 0.863391 0.141636
+vt 0.862626 0.141085
+vt 0.883840 0.136603
+vt 0.883310 0.137383
+vt 0.897417 0.140057
+vt 0.896867 0.140740
+vt 0.898025 0.140801
+vt 0.897174 0.141610
+vt 0.874217 0.133293
+vt 0.874968 0.133747
+vt 0.874208 0.132371
+vt 0.875284 0.132840
+vt 0.896006 0.140567
+vt 0.896002 0.141529
+vt 0.873468 0.133750
+vt 0.873134 0.132848
+vt 0.895621 0.117396
+vt 0.896476 0.117194
+vt 0.895584 0.116434
+vt 0.896753 0.116313
+vt 0.881903 0.155553
+vt 0.881031 0.155656
+vt 0.882465 0.156286
+vt 0.881325 0.156572
+vt 0.897048 0.117859
+vt 0.897631 0.117095
+vt 0.882214 0.154732
+vt 0.883023 0.155251
+vt 0.882691 0.123051
+vt 0.882773 0.122155
+vt 0.881783 0.122789
+vt 0.882031 0.121638
+vt 0.867329 0.155578
+vt 0.866946 0.154764
+vt 0.866810 0.156317
+vt 0.866182 0.155321
+vt 0.883582 0.121769
+vt 0.883025 0.121005
+vt 0.868222 0.155664
+vt 0.867960 0.156572
+vt 0.904030 0.128049
+vt 0.904478 0.128840
+vt 0.904906 0.127731
+vt 0.905368 0.128825
+vt 0.885439 0.141872
+vt 0.885289 0.142769
+vt 0.886281 0.141582
+vt 0.886221 0.142768
+vt 0.904056 0.129646
+vt 0.904943 0.129934
+vt 0.884767 0.141260
+vt 0.885502 0.140686
+vt 0.865721 0.149838
+vt 0.865007 0.147749
+vt 0.865827 0.147457
+vt 0.866565 0.149562
+vt 0.883268 0.130306
+vt 0.883225 0.128099
+vt 0.884112 0.128070
+vt 0.884138 0.130300
+vn 0.0000 -1.0000 0.0000
+vn 0.1760 -0.1821 0.9674
+vn 0.1760 0.1821 0.9674
+vn 0.1373 0.1995 0.9702
+vn 0.1373 -0.1995 0.9702
+vn -0.0000 1.0000 -0.0000
+vn 0.8706 -0.1992 0.4499
+vn 0.8706 0.1992 0.4499
+vn 0.8522 0.1751 0.4930
+vn 0.8522 -0.1751 0.4930
+vn -0.9656 -0.1964 0.1706
+vn -0.9656 0.1964 0.1706
+vn -0.9767 0.1745 0.1252
+vn -0.9767 -0.1745 0.1252
+vn -0.4160 0.1947 -0.8883
+vn -0.4160 -0.1947 -0.8883
+vn -0.9185 -0.2428 -0.3119
+vn -0.9185 0.2428 -0.3119
+vn 0.9185 -0.2429 -0.3119
+vn 0.9185 0.2429 -0.3119
+vn -0.1372 -0.1993 0.9703
+vn -0.1372 0.1993 0.9703
+vn -0.1761 0.1819 0.9674
+vn -0.1761 -0.1819 0.9674
+vn 0.9766 -0.1749 0.1252
+vn 0.9766 0.1749 0.1252
+vn 0.9655 0.1968 0.1705
+vn 0.9655 -0.1968 0.1705
+vn -0.8523 -0.1748 0.4930
+vn -0.8523 0.1748 0.4930
+vn 0.7051 0.1756 -0.6870
+vn 0.7051 -0.1756 -0.6870
+vn 0.6736 -0.1964 -0.7125
+vn 0.6736 0.1964 -0.7125
+vn 0.4161 0.1947 -0.8883
+vn 0.4161 -0.1947 -0.8883
+vn 0.3770 -0.1709 -0.9103
+vn 0.3770 0.1709 -0.9103
+vn -0.8706 -0.1990 0.4499
+vn -0.8706 0.1990 0.4499
+vn 0.5621 0.2389 0.7918
+vn 0.5621 -0.2389 0.7918
+vn -0.5620 0.2388 0.7919
+vn -0.5620 -0.2388 0.7919
+vn -0.7052 -0.1752 -0.6870
+vn -0.7052 0.1752 -0.6870
+vn -0.6736 0.1961 -0.7126
+vn -0.6736 -0.1961 -0.7126
+vn -0.3770 -0.1709 -0.9103
+vn -0.3770 0.1709 -0.9103
+vn 0.0001 -0.2357 -0.9718
+vn 0.0001 0.2357 -0.9718
+vn 0.0822 -0.9807 -0.1776
+vn 0.1339 -0.9813 -0.1382
+vn 0.5148 -0.6952 -0.5016
+vn 0.2776 -0.6869 -0.6716
+vn 0.1339 0.9813 -0.1382
+vn 0.0822 0.9807 -0.1776
+vn 0.2776 0.6869 -0.6716
+vn 0.5148 0.6952 -0.5016
+vn -0.0822 0.9807 -0.1777
+vn -0.1341 0.9812 -0.1385
+vn -0.5153 0.6946 -0.5020
+vn -0.2777 0.6868 -0.6717
+vn -0.1341 -0.9812 -0.1385
+vn -0.0822 -0.9807 -0.1777
+vn -0.2777 -0.6868 -0.6717
+vn -0.5153 -0.6946 -0.5020
+vn 0.1701 0.9813 0.0905
+vn 0.1906 0.9812 0.0307
+vn 0.7146 0.6935 0.0914
+vn 0.6235 0.6939 0.3603
+vn 0.1906 -0.9812 0.0307
+vn 0.1701 -0.9813 0.0905
+vn 0.6235 -0.6939 0.3603
+vn 0.7146 -0.6935 0.0914
+vn -0.1911 0.9811 0.0308
+vn -0.1703 0.9812 0.0907
+vn -0.6239 0.6934 0.3605
+vn -0.7153 0.6928 0.0914
+vn -0.1703 -0.9812 0.0907
+vn -0.1911 -0.9811 0.0308
+vn -0.7153 -0.6928 0.0914
+vn -0.6239 -0.6934 0.3605
+vn -0.1140 -0.9807 0.1590
+vn 0.0574 -0.9732 0.2225
+vn 0.1230 -0.6920 0.7113
+vn -0.3997 -0.7242 0.5619
+vn 0.0574 0.9732 0.2225
+vn -0.1140 0.9807 0.1590
+vn -0.3997 0.7242 0.5619
+vn 0.1230 0.6920 0.7113
+vn -0.2413 -0.9704 0.0099
+vn -0.7222 -0.6844 0.1004
+vn -0.2413 0.9704 0.0099
+vn -0.7222 0.6844 0.1004
+vn -0.1821 0.9811 -0.0651
+vn -0.0726 0.9686 -0.2377
+vn -0.2844 0.6806 -0.6752
+vn -0.6483 0.7279 -0.2233
+vn -0.0726 -0.9686 -0.2377
+vn -0.1821 -0.9811 -0.0651
+vn -0.6483 -0.7279 -0.2233
+vn -0.2844 -0.6806 -0.6752
+vn -0.1966 0.9707 0.1379
+vn -0.6321 0.6875 0.3576
+vn -0.1966 -0.9707 0.1379
+vn -0.6321 -0.6875 0.3576
+vn 0.1820 0.9811 -0.0652
+vn 0.1964 0.9708 0.1377
+vn 0.6318 0.6878 0.3574
+vn 0.6481 0.7280 -0.2234
+vn 0.1964 -0.9708 0.1377
+vn 0.1820 -0.9811 -0.0652
+vn 0.6481 -0.7280 -0.2234
+vn 0.6318 -0.6878 0.3574
+vn 0.0726 0.9686 -0.2378
+vn 0.2843 0.6806 -0.6752
+vn 0.0726 -0.9686 -0.2378
+vn 0.2843 -0.6806 -0.6752
+vn 0.1137 0.9807 0.1592
+vn -0.0576 0.9731 0.2229
+vn -0.1231 0.6917 0.7116
+vn 0.3994 0.7243 0.5620
+vn -0.0576 -0.9731 0.2229
+vn 0.1137 -0.9807 0.1592
+vn 0.3994 -0.7243 0.5620
+vn -0.1231 -0.6917 0.7116
+vn 0.2406 0.9706 0.0100
+vn 0.7217 0.6849 0.1003
+vn 0.2407 -0.9706 0.0100
+vn 0.7217 -0.6849 0.1003
+vn -0.0002 0.9803 -0.1975
+vn 0.1845 0.9711 -0.1512
+vn 0.5181 0.6852 -0.5119
+vn -0.0002 0.7214 -0.6926
+vn 0.1845 -0.9711 -0.1512
+vn -0.0002 -0.9803 -0.1975
+vn -0.0002 -0.7214 -0.6926
+vn 0.5181 -0.6852 -0.5119
+vn -0.1850 0.9710 -0.1515
+vn -0.5185 0.6847 -0.5122
+vn -0.1850 -0.9710 -0.1515
+vn -0.5185 -0.6847 -0.5122
+vn 0.1263 -0.7034 0.6995
+vn -0.1265 -0.7031 0.6997
+vn 0.0277 -0.9822 0.1859
+vn -0.0277 -0.9821 0.1861
+vn -0.1265 0.7031 0.6997
+vn 0.1263 0.7034 0.6995
+vn -0.0277 0.9821 0.1861
+vn 0.0277 0.9822 0.1859
+usemtl toybox.002
+s 1
+f 136/1/1 3/2/1 1/3/1
+f 31/4/1 69/5/1 1/3/1
+f 144/6/2 148/7/3 63/8/4 52/9/5
+f 96/10/6 103/11/6 2/12/6
+f 7/13/1 105/14/1 1/3/1
+f 19/15/6 78/16/6 2/12/6
+f 43/17/1 87/18/1 1/3/1
+f 127/19/6 15/20/6 2/12/6
+f 15/20/6 19/15/6 2/12/6
+f 108/21/7 115/22/8 48/23/9 40/24/10
+f 54/25/11 61/26/12 36/27/13 28/28/14
+f 73/29/1 21/30/1 1/3/1
+f 109/31/1 39/32/1 1/3/1
+f 1/3/1 145/33/1 51/34/1
+f 60/35/6 66/36/6 2/12/6
+f 75/37/1 73/29/1 1/3/1
+f 79/38/15 72/39/16 76/40/17 84/41/18
+f 111/42/1 109/31/1 1/3/1
+f 115/22/8 108/21/7 112/43/19 120/44/20
+f 93/45/1 91/46/1 1/3/1
+f 90/47/21 97/48/22 150/49/23 142/50/24
+f 130/51/6 127/19/6 2/12/6
+f 132/52/1 138/53/1 1/3/1
+f 55/54/1 27/55/1 1/3/1
+f 42/56/25 46/57/26 99/58/27 88/59/28
+f 78/16/6 85/60/6 2/12/6
+f 91/46/1 141/61/1 1/3/1
+f 69/5/1 75/37/1 1/3/1
+f 30/62/29 28/28/14 36/27/13 34/63/30
+f 12/64/31 4/65/32 135/66/33 124/67/34
+f 117/68/35 106/69/36 6/70/37 10/71/38
+f 84/41/18 76/40/17 70/72/39 81/73/40
+f 33/74/6 37/75/6 2/12/6
+f 42/56/25 40/24/10 48/23/9 46/57/26
+f 102/76/41 94/77/42 88/59/28 99/58/27
+f 31/4/1 1/3/1 27/55/1
+f 52/9/5 63/8/4 67/78/43 57/79/44
+f 25/80/1 132/52/1 1/3/1
+f 118/81/6 9/82/6 2/12/6
+f 114/83/6 121/84/6 2/12/6
+f 58/85/1 55/54/1 1/3/1
+f 148/7/3 144/6/2 142/50/24 150/49/23
+f 57/79/44 67/78/43 61/26/12 54/25/11
+f 45/86/6 49/87/6 2/12/6
+f 120/44/20 112/43/19 106/69/36 117/68/35
+f 138/53/1 136/1/1 1/3/1
+f 9/82/6 13/88/6 2/12/6
+f 30/62/29 34/63/30 81/73/40 70/72/39
+f 39/32/1 43/17/1 1/3/1
+f 51/34/1 58/85/1 1/3/1
+f 24/89/45 16/90/46 126/91/47 133/92/48
+f 24/89/45 22/93/49 18/94/50 16/90/46
+f 87/18/1 93/45/1 1/3/1
+f 103/11/6 100/95/6 2/12/6
+f 105/14/1 111/42/1 1/3/1
+f 82/96/6 33/74/6 2/12/6
+f 37/75/6 60/35/6 2/12/6
+f 72/39/16 79/38/15 18/94/50 22/93/49
+f 121/84/6 118/81/6 2/12/6
+f 100/95/6 45/86/6 2/12/6
+f 124/67/34 135/66/33 139/97/51 129/98/52
+f 25/80/1 1/3/1 21/30/1
+f 64/99/6 147/100/6 2/12/6
+f 123/101/6 130/51/6 2/12/6
+f 66/36/6 64/99/6 2/12/6
+f 129/102/52 139/103/51 133/92/48 126/91/47
+f 49/87/6 114/83/6 2/12/6
+f 13/88/6 123/101/6 2/12/6
+f 85/60/6 82/96/6 2/12/6
+f 2/12/6 151/104/6 96/10/6
+f 3/2/1 7/13/1 1/3/1
+f 147/100/6 151/104/6 2/12/6
+f 141/61/1 145/33/1 1/3/1
+f 97/48/22 90/47/21 94/77/42 102/76/41
+f 7/13/53 3/2/54 5/105/55 8/106/56
+f 8/106/56 5/105/55 4/107/32 6/108/37
+f 13/88/57 9/82/58 11/109/59 14/110/60
+f 14/110/60 11/109/59 10/111/38 12/112/31
+f 19/15/61 15/20/62 17/113/63 20/114/64
+f 20/114/64 17/113/63 16/115/46 18/116/50
+f 25/80/65 21/30/66 23/117/67 26/118/68
+f 26/118/68 23/117/67 22/119/49 24/120/45
+f 49/87/69 45/86/70 47/121/71 50/122/72
+f 50/122/72 47/121/71 46/123/26 48/124/9
+f 43/17/73 39/32/74 41/125/75 44/126/76
+f 44/126/76 41/125/75 40/127/10 42/128/25
+f 37/75/77 33/74/78 35/129/79 38/130/80
+f 38/130/80 35/129/79 34/131/30 36/132/13
+f 31/4/81 27/55/82 29/133/83 32/134/84
+f 32/134/84 29/133/83 28/135/14 30/136/29
+f 58/85/85 51/34/86 53/137/87 59/138/88
+f 59/138/88 53/137/87 52/139/5 57/140/44
+f 64/99/89 66/36/90 68/141/91 65/142/92
+f 65/142/92 68/141/91 67/143/43 63/144/4
+f 55/54/93 58/85/85 59/138/88 56/145/94
+f 56/145/94 59/138/88 57/140/44 54/146/11
+f 66/36/90 60/35/95 62/147/96 68/141/91
+f 68/141/91 62/147/96 61/148/12 67/143/43
+f 85/60/97 78/16/98 80/149/99 86/150/100
+f 86/150/100 80/149/99 79/151/15 84/152/18
+f 73/29/101 75/37/102 77/153/103 74/154/104
+f 74/154/104 77/153/103 76/155/17 72/156/16
+f 82/96/105 85/60/97 86/150/100 83/157/106
+f 83/157/106 86/150/100 84/152/18 81/158/40
+f 75/37/102 69/5/107 71/159/108 77/153/103
+f 77/153/103 71/159/108 70/160/39 76/155/17
+f 121/84/109 114/83/110 116/161/111 122/162/112
+f 122/162/112 116/161/111 115/163/8 120/164/20
+f 109/31/113 111/42/114 113/165/115 110/166/116
+f 110/166/116 113/165/115 112/167/19 108/168/7
+f 118/81/117 121/84/109 122/162/112 119/169/118
+f 119/169/118 122/162/112 120/164/20 117/170/35
+f 111/42/114 105/14/119 107/171/120 113/165/115
+f 113/165/115 107/171/120 106/172/36 112/167/19
+f 103/11/121 96/10/122 98/173/123 104/174/124
+f 104/174/124 98/173/123 97/175/22 102/176/41
+f 91/46/125 93/45/126 95/177/127 92/178/128
+f 92/178/128 95/177/127 94/179/42 90/180/21
+f 100/95/129 103/11/121 104/174/124 101/181/130
+f 101/181/130 104/174/124 102/176/41 99/182/27
+f 93/45/126 87/18/131 89/183/132 95/177/127
+f 95/177/127 89/183/132 88/184/28 94/179/42
+f 130/51/133 123/101/134 125/185/135 131/186/136
+f 131/186/136 125/185/135 124/187/34 129/188/52
+f 136/1/137 138/53/138 140/189/139 137/190/140
+f 137/190/140 140/189/139 139/191/51 135/192/33
+f 127/19/141 130/51/133 131/186/136 128/193/142
+f 128/193/142 131/186/136 129/188/52 126/194/47
+f 138/53/138 132/52/143 134/195/144 140/189/139
+f 140/189/139 134/195/144 133/196/48 139/191/51
+f 27/55/82 55/54/93 56/145/94 29/133/83
+f 29/133/83 56/145/94 54/146/11 28/135/14
+f 60/35/95 37/75/77 38/130/80 62/147/96
+f 62/147/96 38/130/80 36/132/13 61/148/12
+f 69/5/107 31/4/81 32/134/84 71/159/108
+f 71/159/108 32/134/84 30/136/29 70/160/39
+f 33/74/78 82/96/105 83/157/106 35/129/79
+f 35/129/79 83/157/106 81/158/40 34/131/30
+f 21/30/66 73/29/101 74/154/104 23/117/67
+f 23/117/67 74/154/104 72/156/16 22/119/49
+f 78/16/98 19/15/61 20/114/64 80/149/99
+f 80/149/99 20/114/64 18/116/50 79/151/15
+f 87/18/131 43/17/73 44/126/76 89/183/132
+f 89/183/132 44/126/76 42/128/25 88/184/28
+f 45/86/70 100/95/129 101/181/130 47/121/71
+f 47/121/71 101/181/130 99/182/27 46/123/26
+f 9/82/58 118/81/117 119/169/118 11/109/59
+f 11/109/59 119/169/118 117/170/35 10/111/38
+f 105/14/119 7/13/53 8/106/56 107/171/120
+f 107/171/120 8/106/56 6/108/37 106/172/36
+f 39/32/74 109/31/113 110/166/116 41/125/75
+f 41/125/75 110/166/116 108/168/7 40/127/10
+f 114/83/110 49/87/69 50/122/72 116/161/111
+f 116/161/111 50/122/72 48/124/9 115/163/8
+f 123/101/134 13/88/57 14/110/60 125/185/135
+f 125/185/135 14/110/60 12/112/31 124/187/34
+f 3/2/54 136/1/137 137/190/140 5/105/55
+f 5/105/55 137/190/140 135/192/33 4/107/32
+f 132/52/143 25/80/65 26/118/68 134/195/144
+f 134/195/144 26/118/68 24/120/45 133/196/48
+f 15/20/62 127/19/141 128/193/142 17/113/63
+f 17/113/63 128/193/142 126/194/47 16/115/46
+f 142/197/24 144/198/2 146/199/145 143/200/146
+f 143/200/146 146/199/145 145/33/147 141/61/148
+f 148/201/3 150/202/23 152/203/149 149/204/150
+f 149/204/150 152/203/149 151/104/151 147/100/152
+f 141/61/148 91/46/125 92/178/128 143/200/146
+f 143/200/146 92/178/128 90/180/21 142/197/24
+f 96/10/122 151/104/151 152/203/149 98/173/123
+f 98/173/123 152/203/149 150/202/23 97/175/22
+f 51/34/86 145/33/147 146/199/145 53/137/87
+f 53/137/87 146/199/145 144/198/2 52/139/5
+f 147/100/152 64/99/89 65/142/92 149/204/150
+f 149/204/150 65/142/92 63/144/4 148/201/3
+f 12/64/31 10/71/38 6/70/37 4/65/32

envs/assets/blocks/yellow_pentagon.urdf

@@ -0,0 +1,30 @@
+<?xml version="1.0" ?>
+<robot name="yellow_pentagon.urdf">
+  <link name="baseLink">
+    <contact>
+      <lateral_friction value="0.5"/>
+      <rolling_friction value="0.0001"/>
+      <inertia_scaling value="1.0"/>
+    </contact>
+    <inertial>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+       <mass value=".01"/>
+       <inertia ixx="1" ixy="0" ixz="0" iyy="1" iyz="0" izz="1"/>
+    </inertial>
+    <visual>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <mesh filename="pentagon.obj" scale="1.0 1.0 1.0"/>
+      </geometry>
+       <material name="yellow">
+        <color rgba="1 1 0.4 1"/>
+      </material>
+    </visual>
+    <collision>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <mesh filename="pentagon.obj" scale="1.0 1.0 1.0"/>
+      </geometry>
+    </collision>
+  </link>
+</robot>

envs/assets/insert.urdf

@@ -0,0 +1,66 @@
+<?xml version="0.0" ?>
+<robot name="ell.urdf">
+  <link name="baseLink">
+    <contact>
+      <lateral_friction value="0.3"/>
+      <rolling_friction value="0.0001"/>
+      <inertia_scaling value="3.0"/>
+    </contact>
+    <inertial>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+       <mass value=".1"/>
+       <inertia ixx="1" ixy="0" ixz="0" iyy="1" iyz="0" izz="1"/>
+    </inertial>
+
+    <visual>
+      <origin rpy="0 0 0" xyz="0.025 0 0"/>
+      <geometry>
+        <box size=".08 .03 .04"/>
+      </geometry>
+      <material name="red">
+        <color rgba="0. 0.3412 0.3490 1"/>
+      </material>
+    </visual>
+    <collision>
+      <origin rpy="0 0 0" xyz="0.025 0 0"/>
+      <geometry>
+        <box size=".08 .03 .04"/>
+      </geometry>
+    </collision>
+
+    <visual>
+      <origin rpy="0 0 0" xyz="0 0.025 0"/>
+      <geometry>
+        <box size=".03 .08 .0399"/>
+      </geometry>
+      <material name="dark">
+        <color rgba="0. 0.3412 0.3490 1"/>
+      </material>
+    </visual>
+    <collision>
+      <origin rpy="0 0 0" xyz="0 0.025 0"/>
+      <geometry>
+        <box size=".03 .08 .0399"/>
+      </geometry>
+    </collision>
+
+
+    <visual>
+      <origin rpy="0 0 0" xyz="0.075 0.025 0"/>
+      <geometry>
+        <box size=".03 .08 .0399"/>
+      </geometry>
+      <material name="dark">
+        <color rgba="0. 0.3412 0.3490 1"/>
+      </material>
+    </visual>
+    <collision>
+      <origin rpy="0 0 0" xyz="0.075 0.025 0"/>
+      <geometry>
+        <box size=".03 .08 .0399"/>
+      </geometry>
+    </collision>
+
+
+  </link>
+</robot>

envs/assets/plane.obj

@@ -0,0 +1,18 @@
+# Blender v2.66 (sub 1) OBJ File: ''
+# www.blender.org
+mtllib plane.mtl
+o Plane
+v 15.000000 -15.000000 0.000000
+v 15.000000 15.000000 0.000000
+v -15.000000 15.000000 0.000000
+v -15.000000  -15.000000 0.000000
+
+vt 15.000000 0.000000
+vt 15.000000 15.000000
+vt 0.000000 15.000000
+vt 0.000000 0.000000
+
+usemtl Material
+s off
+f 1/1 2/2 3/3
+f 1/1 3/3 4/4

envs/assets/suction/base.obj

@@ -0,0 +1,396 @@
+# Object Export From Tinkercad Server 2015
+
+mtllib obj.mtl
+
+o obj_0
+v 7.413         37.27         25
+v 7.413         37.27         0
+v 33.513         17.913         25
+v 35.107         14.542         25
+v 35.107         14.542         0
+v 33.513         17.913         0
+v -31.596         21.112         25
+v -31.596         21.112         0
+v -33.513         17.913         0
+v -33.513         17.913         25
+v 3.725         37.817         0
+v 3.725         37.817         25
+v -29.374         24.107         25
+v -29.374         24.107         0
+v 0         38         0
+v 11.031         36.364         25
+v 11.031         36.364         0
+v 14.542         35.107         25
+v 14.542         35.107         0
+v -37.27         -7.413         25
+v 17.913         33.513         25
+v -37.27         -7.413         0
+v -11.031         -36.364         25
+v 17.913         33.513         0
+v -36.364         -11.031         0
+v -14.542         -35.107         25
+v -26.87         26.87         25
+v -36.364         -11.031         25
+v -14.542         -35.107         0
+v -26.87         26.87         0
+v -3.725         37.817         0
+v -11.031         -36.364         0
+v 24.107         -29.374         25
+v -17.913         -33.513         25
+v -17.913         -33.513         0
+v 21.112         -31.596         25
+v 21.112         -31.596         0
+v 24.107         -29.374         0
+v -3.725         37.817         25
+v 0         38         25
+v 17.913         -33.513         25
+v 17.913         -33.513         0
+v 37.817         -3.725         25
+v -37.817         -3.725         25
+v -37.817         -3.725         0
+v 37.817         -3.725         0
+v 38         0         0
+v -21.112         -31.596         25
+v 38         0         25
+v 14.542         -35.107         25
+v -21.112         -31.596         0
+v 36.364         11.031         25
+v 14.542         -35.107         0
+v 36.364         11.031         0
+v 37.27         -7.413         25
+v 37.27         -7.413         0
+v -24.107         29.374         25
+v 21.112         31.596         25
+v -24.107         29.374         0
+v 36.364         -11.031         25
+v 21.112         31.596         0
+v 36.364         -11.031         0
+v -38         0         25
+v 11.031         -36.364         25
+v 11.031         -36.364         0
+v -21.112         31.596         25
+v -21.112         31.596         0
+v 7.413         -37.27         25
+v 7.413         -37.27         0
+v -17.913         33.513         25
+v -17.913         33.513         0
+v 24.107         29.374         0
+v -38         0         0
+v 26.87         26.87         0
+v 35.107         -14.542         25
+v 29.374         24.107         0
+v 35.107         -14.542         0
+v 31.596         21.112         0
+v 3.725         -37.817         25
+v 37.27         7.413         25
+v 3.725         -37.817         0
+v 24.107         29.374         25
+v -24.107         -29.374         25
+v -14.542         35.107         25
+v -24.107         -29.374         0
+v -14.542         35.107         0
+v 37.27         7.413         0
+v 33.513         -17.913         25
+v 37.817         3.725         0
+v 33.513         -17.913         0
+v -11.031         36.364         25
+v -11.031         36.364         0
+v -7.413         37.27         25
+v 26.87         26.87         25
+v -7.413         37.27         0
+v -26.87         -26.87         25
+v -26.87         -26.87         0
+v 37.817         3.725         25
+v 0         -38         25
+v 31.596         -21.112         25
+v 0         -38         0
+v 31.596         -21.112         0
+v 29.374         24.107         25
+v -29.374         -24.107         25
+v -29.374         -24.107         0
+v 29.374         -24.107         25
+v 31.596         21.112         25
+v 29.374         -24.107         0
+v -3.725         -37.817         25
+v -31.596         -21.112         25
+v -3.725         -37.817         0
+v -31.596         -21.112         0
+v -37.817         3.725         25
+v -37.817         3.725         0
+v 26.87         -26.87         25
+v 26.87         -26.87         0
+v -37.27         7.413         0
+v -37.27         7.413         25
+v -7.413         -37.27         25
+v -33.513         -17.913         25
+v -33.513         -17.913         0
+v -7.413         -37.27         0
+v -36.364         11.031         25
+v -36.364         11.031         0
+v -35.107         -14.542         25
+v -35.107         -14.542         0
+v -35.107         14.542         25
+v -35.107         14.542         0
+# 128 vertices
+
+g group_0_2829873
+
+usemtl color_2829873
+s 0
+
+f 3     4     5
+f 3     5     6
+f 7     8     9
+f 7     9     10
+f 1     2     11
+f 1     11     12
+f 16     17     2
+f 16     2     1
+f 13     14     8
+f 13     8     7
+f 18     19     17
+f 18     17     16
+f 23     26     29
+f 23     29     32
+f 26     34     35
+f 26     35     29
+f 33     36     37
+f 33     37     38
+f 15     31     39
+f 15     39     40
+f 20     22     25
+f 20     25     28
+f 36     41     42
+f 36     42     37
+f 44     45     22
+f 44     22     20
+f 21     24     19
+f 21     19     18
+f 43     46     47
+f 43     47     49
+f 27     30     14
+f 27     14     13
+f 41     50     53
+f 41     53     42
+f 43     55     56
+f 43     56     46
+f 34     48     51
+f 34     51     35
+f 4     52     54
+f 4     54     5
+f 50     64     65
+f 50     65     53
+f 57     59     30
+f 57     30     27
+f 55     60     62
+f 55     62     56
+f 66     67     59
+f 66     59     57
+f 64     68     69
+f 64     69     65
+f 58     61     24
+f 58     24     21
+f 63     73     45
+f 63     45     44
+f 70     71     67
+f 70     67     66
+f 60     75     77
+f 60     77     62
+f 82     72     61
+f 82     61     58
+f 68     79     81
+f 68     81     69
+f 48     83     85
+f 48     85     51
+f 84     86     71
+f 84     71     70
+f 75     88     90
+f 75     90     77
+f 91     92     86
+f 91     86     84
+f 15     11     2
+f 15     2     17
+f 15     17     19
+f 15     19     24
+f 15     24     61
+f 15     61     72
+f 15     72     74
+f 15     74     76
+f 15     76     78
+f 15     78     6
+f 15     6     5
+f 15     5     54
+f 15     54     87
+f 15     87     89
+f 15     89     47
+f 52     80     87
+f 52     87     54
+f 94     74     72
+f 94     72     82
+f 93     95     92
+f 93     92     91
+f 83     96     97
+f 83     97     85
+f 80     98     89
+f 80     89     87
+f 79     99     101
+f 79     101     81
+f 88     100     102
+f 88     102     90
+f 94     103     76
+f 94     76     74
+f 39     31     95
+f 39     95     93
+f 104     105     97
+f 104     97     96
+f 98     49     47
+f 98     47     89
+f 100     106     108
+f 100     108     102
+f 103     107     78
+f 103     78     76
+f 101     99     109
+f 101     109     111
+f 110     112     105
+f 110     105     104
+f 113     114     73
+f 113     73     63
+f 106     115     116
+f 106     116     108
+f 107     3     6
+f 107     6     78
+f 118     117     114
+f 118     114     113
+f 120     121     112
+f 120     112     110
+f 109     119     122
+f 109     122     111
+f 123     124     117
+f 123     117     118
+f 125     126     121
+f 125     121     120
+f 127     128     124
+f 127     124     123
+f 115     33     38
+f 115     38     116
+f 10     9     128
+f 10     128     127
+f 28     25     126
+f 28     126     125
+f 119     23     32
+f 119     32     122
+f 73     114     117
+f 73     117     124
+f 73     124     128
+f 73     128     9
+f 73     9     8
+f 73     8     14
+f 73     14     30
+f 73     30     59
+f 73     59     67
+f 73     67     71
+f 73     71     86
+f 73     86     92
+f 73     92     95
+f 73     95     31
+f 73     31     15
+f 81     101     15
+f 47     46     15
+f 46     56     15
+f 56     62     15
+f 62     77     15
+f 77     90     15
+f 90     102     15
+f 102     108     15
+f 108     116     15
+f 116     38     15
+f 38     37     15
+f 37     42     15
+f 42     53     15
+f 53     65     15
+f 69     15     65
+f 81     15     69
+f 45     73     15
+f 101     111     15
+f 111     122     15
+f 122     32     15
+f 32     29     15
+f 29     35     15
+f 35     51     15
+f 51     85     15
+f 85     97     15
+f 97     105     15
+f 105     112     15
+f 112     121     15
+f 121     126     15
+f 126     25     15
+f 22     15     25
+f 45     15     22
+f 49     98     80
+f 49     80     52
+f 49     52     4
+f 49     4     3
+f 49     3     107
+f 49     107     103
+f 49     103     94
+f 49     94     82
+f 49     82     58
+f 49     58     21
+f 49     21     18
+f 49     18     16
+f 49     16     1
+f 49     1     12
+f 49     12     40
+f 40     39     93
+f 40     93     91
+f 40     91     84
+f 40     84     70
+f 40     70     66
+f 40     66     57
+f 40     57     27
+f 40     27     13
+f 40     13     7
+f 40     7     10
+f 40     10     127
+f 40     127     123
+f 40     123     118
+f 40     118     113
+f 63     40     113
+f 68     40     79
+f 64     40     68
+f 50     40     64
+f 41     40     50
+f 36     40     41
+f 33     40     36
+f 115     40     33
+f 106     40     115
+f 100     40     106
+f 88     40     100
+f 75     40     88
+f 60     40     75
+f 55     40     60
+f 43     40     55
+f 49     40     43
+f 99     79     40
+f 20     40     44
+f 28     40     20
+f 125     40     28
+f 120     40     125
+f 110     40     120
+f 104     40     110
+f 96     40     104
+f 83     40     96
+f 48     40     83
+f 34     40     48
+f 26     40     34
+f 23     40     26
+f 119     40     23
+f 109     40     119
+f 99     40     109
+f 63     44     40
+f 12     11     15
+f 12     15     40
+# 252 faces
+
+ #end of obj_0
+

envs/assets/suction/cylinder.urdf

@@ -0,0 +1,98 @@
+<?xml version="0.0" ?>
+<robot name="cylinder.urdf">
+  <link name="headLink">
+    <contact>
+      <lateral_friction value="1.0"/>
+      <rolling_friction value="0.0001"/>
+      <inertia_scaling value="3.0"/>
+    </contact>
+    <inertial>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+       <mass value=".1"/>
+       <inertia ixx="1" ixy="0" ixz="0" iyy="1" iyz="0" izz="1"/>
+    </inertial>
+    <visual>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <mesh filename="head.obj" scale="0.001 0.001 0.001"/>
+      </geometry>
+      <material name="darkgrey">
+        <color rgba="0.2 0.2 0.2 1"/>
+      </material>
+    </visual>
+    <collision>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <mesh filename="head.obj" scale="0.001 0.001 0.001"/>
+      </geometry>
+    </collision>
+  </link>
+
+  <joint name="tipJoint" type="fixed">
+    <parent link="headLink"/>
+    <child link="tipLink"/>
+    <origin rpy="0.0 0.0 0.0" xyz="0.0 0.0 0.029"/>
+    <axis xyz="0 0 1"/>
+    <limit effort="150.0" lower="-6.28318530718" upper="6.28318530718" velocity="3.15"/>
+    <dynamics damping="10.0" friction="0.0"/>
+  </joint>
+
+  <link name="tipLink">
+    <contact>
+      <lateral_friction value="1.0"/>
+      <rolling_friction value="0.0001"/>
+      <inertia_scaling value="3.0"/>
+    </contact>
+    <inertial>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+       <mass value=".1"/>
+       <inertia ixx="1" ixy="0" ixz="0" iyy="1" iyz="0" izz="1"/>
+    </inertial>
+    <visual>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <cylinder length="0.05" radius="0.005"/>
+      </geometry>
+      <material name="blue">
+        <color rgba="0.18039216, 0.50588235, 0.77254902 1"/>
+      </material>
+    </visual>
+    <collision>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <cylinder length="0.05" radius="0.005"/>
+      </geometry>
+    </collision>
+  </link>
+
+  <!-- <link name="asdfLink">
+    <contact>
+      <lateral_friction value="1.0"/>
+      <rolling_friction value="0.0001"/>
+      <inertia_scaling value="3.0"/>
+    </contact>
+    <inertial>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+       <mass value=".1"/>
+       <inertia ixx="1" ixy="0" ixz="0" iyy="1" iyz="0" izz="1"/>
+    </inertial>
+    <visual>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <cylinder length="0.028" radius="0.001"/>
+      </geometry>
+    </visual>
+  </link>
+
+  <joint name="asdfoint" type="fixed">
+    <parent link="tipLink"/>
+    <child link="asdfLink"/>
+    <origin rpy="0.0 0.0 0.0" xyz="0.0 0.0 0.0"/>
+    <axis xyz="0 0 1"/>
+    <limit effort="150.0" lower="-6.28318530718" upper="6.28318530718" velocity="3.15"/>
+    <dynamics damping="10.0" friction="0.0"/>
+  </joint> -->
+
+
+</robot>
+

envs/assets/suction/cylinder_real.urdf

@@ -0,0 +1,98 @@
+<?xml version="0.0" ?>
+<robot name="cylinder_real.urdf">
+  <link name="headLink">
+    <contact>
+      <lateral_friction value="1.0"/>
+      <rolling_friction value="0.0001"/>
+      <inertia_scaling value="3.0"/>
+    </contact>
+    <inertial>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+       <mass value=".1"/>
+       <inertia ixx="1" ixy="0" ixz="0" iyy="1" iyz="0" izz="1"/>
+    </inertial>
+    <visual>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <mesh filename="head.obj" scale="0.001 0.001 0.001"/>
+      </geometry>
+      <material name="darkgrey">
+        <color rgba="0.2 0.2 0.2 1"/>
+      </material>
+    </visual>
+    <collision>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <mesh filename="head.obj" scale="0.001 0.001 0.001"/>
+      </geometry>
+    </collision>
+  </link>
+
+  <joint name="tipJoint" type="fixed">
+    <parent link="headLink"/>
+    <child link="tipLink"/>
+    <origin rpy="0.0 0.0 0.0" xyz="0.0 0.0 0.029"/>
+    <axis xyz="0 0 1"/>
+    <limit effort="150.0" lower="-6.28318530718" upper="6.28318530718" velocity="3.15"/>
+    <dynamics damping="10.0" friction="0.0"/>
+  </joint>
+
+  <link name="tipLink">
+    <contact>
+      <lateral_friction value="1.0"/>
+      <rolling_friction value="0.0001"/>
+      <inertia_scaling value="3.0"/>
+    </contact>
+    <inertial>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+       <mass value=".1"/>
+       <inertia ixx="1" ixy="0" ixz="0" iyy="1" iyz="0" izz="1"/>
+    </inertial>
+    <visual>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <cylinder length="0.135" radius="0.0127"/>
+      </geometry>
+      <material name="blue">
+        <color rgba="0.5, 0.5, 0.5 1"/>
+      </material>
+    </visual>
+    <collision>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <cylinder length="0.135" radius="0.0127"/>
+      </geometry>
+    </collision>
+  </link>
+
+  <!-- <link name="asdfLink">
+    <contact>
+      <lateral_friction value="1.0"/>
+      <rolling_friction value="0.0001"/>
+      <inertia_scaling value="3.0"/>
+    </contact>
+    <inertial>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+       <mass value=".1"/>
+       <inertia ixx="1" ixy="0" ixz="0" iyy="1" iyz="0" izz="1"/>
+    </inertial>
+    <visual>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+        <cylinder length="0.028" radius="0.001"/>
+      </geometry>
+    </visual>
+  </link>
+
+  <joint name="asdfoint" type="fixed">
+    <parent link="tipLink"/>
+    <child link="asdfLink"/>
+    <origin rpy="0.0 0.0 0.0" xyz="0.0 0.0 0.0"/>
+    <axis xyz="0 0 1"/>
+    <limit effort="150.0" lower="-6.28318530718" upper="6.28318530718" velocity="3.15"/>
+    <dynamics damping="10.0" friction="0.0"/>
+  </joint> -->
+
+
+</robot>
+