first commit

.gitignore

@@ -0,0 +1,130 @@
+.vscode/
+
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+pip-wheel-metadata/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+.python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/

.pre-commit-config.yaml

@@ -0,0 +1,59 @@
+minimum_pre_commit_version: 2.20.0
+default_stages: [commit, push]
+exclude: "^$"
+repos:
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v4.4.0
+    hooks:
+      - id: check-case-conflict
+      - id: check-executables-have-shebangs
+      - id: check-merge-conflict
+      - id: check-shebang-scripts-are-executable
+      - id: check-symlinks
+      - id: check-yaml
+      - id: debug-statements
+      - id: destroyed-symlinks
+      - id: end-of-file-fixer
+        files: \.(py|sh|rst|yml|yaml)$
+      - id: mixed-line-ending
+      - id: trailing-whitespace
+        files: \.(py|sh|rst|yml|yaml)$
+  - repo: https://github.com/psf/black
+    rev: 23.7.0
+    hooks:
+      - id: black
+  - repo: https://github.com/PyCQA/flake8
+    rev: 6.0.0
+    hooks:
+      - id: flake8
+        args:
+          - --config=setup.cfg
+          - --ignore=E203,E266,E501,W503,F403,F401,B905
+        additional_dependencies:
+          [
+            "flake8-blind-except",
+            "flake8-docstrings",
+            "flake8-bugbear",
+            "flake8-comprehensions",
+            "flake8-docstrings",
+            "flake8-implicit-str-concat",
+            "pydocstyle>=5.0.0",
+          ]
+        exclude: ^scalingup/(prompts|policy/prompts|task_generator/prompts)/|build/|scripts/|scalingup/algo/robomimic_nets|scalingup/algo/tsdf
+  - repo: https://github.com/pre-commit/mirrors-mypy
+    rev: "v1.4.1"
+    hooks:
+      - id: mypy
+        args:
+          - --check-untyped-defs
+        additional_dependencies:
+          - "pydantic"
+        exclude: ^scalingup/(prompts)/|tests/|build
+  - repo: https://github.com/codespell-project/codespell
+    rev: v2.2.5
+    hooks:
+      - id: codespell
+        args:
+          - -L caler
+          - --write
+        exclude: scripts/visualizations/rollout|presentation/js|presentation/plugin|presentation/package-lock.json|.svg|presentation/index.html|scalingup/algo/tsdf.py

README.md

@@ -0,0 +1,105 @@
+<h1> Scaling Up and Distilling Down: Language-Guided Robot Skill Acquisition</h1>
+<div style="text-align: center;">
+
+[Huy Ha](https://www.cs.columbia.edu/~huy/)$^1$, [Pete Florence](https://www.peteflorence.com/)$^2$,  [Shuran Song](https://shurans.github.io/)$^1$
+
+$^1$ Columbia University, $^2$ Google DeepMind
+
+[Project Page](https://www.cs.columbia.edu/~huy/scalingup) | [Arxiv](https://arxiv.org/abs/2307.14535) | [Video](https://www.cs.columbia.edu/~huy/scalingup/static/videos/scalingup.mp4)
+
+<div style="margin:50px; text-align: justify;">
+<img style="width:100%;" src="docs/assets/teaser.gif">
+
+Scaling Up and Distilling Down is a framework for language-guided skill learning.
+Give it a task description, and it will automatically generate rich, diverse robot trajectories, complete with success label and dense language labels.
+
+<b>The best part?</b> It uses no expert demonstrations, manual
+reward supervision, and no manual language annotation.
+
+</div>
+</div>
+
+<br>
+
+This repository contains code for language-guided data generation and language-conditioned diffusion policy training for [Scaling Up And Distilling Down](https://www.cs.columbia.edu/~huy/scalingup).
+It has been tested on Ubuntu 18.04, 20.04 and 22.04, NVIDIA GTX 1080, NVIDIA RTX A6000, NVIDIA GeForce RTX 3080, and NVIDIA GeForce RTX 3090.
+
+If you find this codebase useful, consider citing:
+
+```bibtex
+@inproceedings{ha2023scalingup,
+      title={Scaling Up and Distilling Down: Language-Guided Robot Skill Acquisition}, 
+      author={Huy Ha and Pete Florence and Shuran Song},
+      year={2023},
+      eprint={2307.14535},
+      archivePrefix={arXiv},
+      primaryClass={cs.RO}
+}
+```
+
+If you have any questions, please contact [me](https://www.cs.columbia.edu/~huy/) at `huy [at] cs [dot] columbia [dot] edu`.
+
+**Table of Contents**
+
+ - ⚙️[Setup](docs/setup.md)
+ - 🚶[Codebase Walkthrough](docs/walkthrough.md)
+   - [💡 Core Concepts](docs/walkthrough.md#-core-concepts)
+      - [🪺 Nested Trajectories using Hierachical Actions and Policies](docs/walkthrough.md#-nested-trajectories-using-hierachical-actions-and-policies)
+      - [🌳 Exploration Task Tree](docs/walkthrough.md#exploration-task-tree)
+      - [🌈 Seeded Variation](docs/walkthrough.md#🌈-seeded-variation)
+    - [🎛️ Control](docs/walkthrough.md#control)
+    - [🏃‍♂️ Motion Planning](docs/walkthrough.md#️-motion-planning)
+    - [🗣️ Language Model Queries](docs/walkthrough.md#language-model-queries)
+      - [💿 Cache](docs/walkthrough.md#-cache)
+      - [🔗 Linking LLM Modules Together](docs/walkthrough.md#-linking-llm-modules-together)
+      - [🪙 Coin flips](docs/walkthrough.md#-coin-flips)
+ - 🔬 [Reproducing](docs/reproduce.md) 
+   - 📊 [Evaluation](docs/reproduce.md#evaluation) 
+   - 🗄️ [Data Generation](docs/reproduce.md#data-generation-for-training)
+   - 🧠 [Training](docs/reproduce.md#training)
+- 🔭 [Extending](docs/extend.md)
+    - [I want to add more](docs/extend.md#i-want-to-add-more)
+      - 🤖 [Robots](docs/extend.md#robots-)
+      - 🪑 [Assets](docs/extend.md#assets-)
+        - 📜 [Tools \& Scripts](docs/extend.md#tools--scripts-)
+      - 🌏 [Environments \& Tasks](docs/extend.md#environments--tasks-)
+        - [New Simulators](docs/extend.md#new-simulators)
+        - [New Tasks](docs/extend.md#new-tasks)
+      - 🦙 [Language Models](docs/extend.md#language-models-)
+    - 🖼️ [Figure Utilities](docs/extend.md#figure-utilities-️)
+      - [Efficiency Plot](docs/extend.md#efficiency-plot)
+      - [Visualizing Language-conditioned Outputs](docs/extend.md#visualizing-language-conditioned-outputs)
+    - ✅ [Development Tips](docs/extend.md#development-tips-)
+        - 🐉 [Hydra](docs/extend.md#hydra-)
+        - 📷 [Headless Rendering](docs/extend.md#headless-rendering-)
+        - 🖧 [Multi-processing](docs/extend.md#multi-processing-)
+        - 👩‍👦‍👶 [Typing](docs/extend.md#typing-)
+        - ⏱️ [Profiling](docs/extend.md#profiling-️)
+        - 💽 [Data Format](docs/extend.md#data-format-)
+        - 💾 [RAM Usage](docs/extend.md#ram-usage-)
+    - 💀 [Known Issues](docs/extend.md#known-issues-)
+    - ✅ [Training Tips](docs/extend.md#training-tips-)
+      - [Mixed-Precision](docs/extend.md#mixed-precision)
+ - 📽️ [Visualizations](docs/visualization.md)
+
+# Acknowledgements
+
+
+We would like to thank Cheng Chi, Zeyi Liu, Samir Yitzhak Gadre, Mengda Xu, Zhenjia Xu, Mandi Zhao and Dominik Bauer for their helpful feedback and fruitful discussions.
+
+
+This work was supported in part by Google Research Award, NSF Award #2143601, and #2132519. 
+We would like to thank Google for the UR5 robot hardware.
+The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies, either expressed or implied, of the sponsors.
+
+## Code
+
+ - [Diffusion Policy](https://diffusion-policy.cs.columbia.edu/): The policy was built on top of their [Colab](https://colab.research.google.com/drive/18GIHeOQ5DyjMN8iIRZL2EKZ0745NLIpg?usp=sharing), and our real-world evaluation code was modified from [theirs](https://github.com/columbia-ai-robotics/diffusion_policy#-demo-training-and-eval-on-a-real-robot)
+ - [Mujoco Menagerie](https://github.com/deepmind/mujoco_menagerie): UR5 and RealSense models were modified from their models.
+ - [Mujoco Scanned Objects](https://github.com/kevinzakka/mujoco_scanned_objects): Big shout out to [Kevin Zakka](https://kzakka.com/) for all his amazing open-source work, go give him a few stars ⭐
+ - [3D UNet Implementation](https://github.com/wolny/pytorch-3dunet/) modified from [Adrian Wolny](https://adrianwolny.com/)'s implementation.
+
+
+## Coming Soon
+
+ - **Data + Checkpoint Downloads**: For now, you can still run data generation and training!