Rapid experimentation and scaling of deep learning models on molecular and crystal graphs.

Overview

LitMatter

A template for rapid experimentation and scaling deep learning models on molecular and crystal graphs.

How to use

  1. Clone this repository and start editing, or save it and use it as a template for new projects.
  2. Edit lit_models/models.py with the PyTorch code for your model of interest.
  3. Edit lit_data/data.py to load and process your PyTorch datasets.
  4. Perform interactive experiments in prototyping.py.
  5. Scale network training to any number of GPUs using the example batch scripts.

Principles

LitMatter uses PyTorch Lightning to organize PyTorch code so scientists can rapidly experiment with geometric deep learning and scale up to hundreds of GPUs without difficulty. Many amazing applied ML methods (even those with open-source code) are never used by the wider community because the important details are buried in hundreds of lines of boilerplate code. It may require a significant engineering effort to get the method working on a new dataset and in a different computing environment, and it can be hard to justify this effort before verifying that the method will provide some advantage. Packaging your code with the LitMatter template makes it easy for other researchers to experiment with your models and scale them beyond common benchmark datasets.

Features

  • Maximum flexibility. LitMatter supports arbitrary PyTorch models and dataloaders.
  • Eliminate boilerplate. Engineering code is abstracted away, but still accessible if needed.
  • Full end-to-end pipeline. Data processing, model construction, training, and inference can be launched from the command line, in a Jupyter notebook, or through a SLURM job.
  • Lightweight. Using the template is easier than not using it; it reduces infrastructure overhead for simple and complex deep learning projects.

Examples

The example notebooks show how to use LitMatter to scale model training for different applications.

  • Prototyping GNNs - train an equivariant graph neural network to predict quantum properties of small molecules.
  • Neural Force Fields - train a neural force field on molecular dynamics trajectories of small molecules.
  • DeepChem - train a PyTorch model in DeepChem on a MoleculeNet dataset.
  • 🤗 - train a 🤗 language model to generate molecules.

Note that these examples have additional dependencies beyond the core depdencies of LitMatter.

References

If you use LitMatter for your own research and scaling experiments, please cite the following work: Frey, Nathan C., et al. "Scalable Geometric Deep Learning on Molecular Graphs." NeurIPS 2021 AI for Science Workshop. 2021.

@inproceedings{frey2021scalable,
  title={Scalable Geometric Deep Learning on Molecular Graphs},
  author={Frey, Nathan C and Samsi, Siddharth and McDonald, Joseph and Li, Lin and Coley, Connor W and Gadepally, Vijay},
  booktitle={NeurIPS 2021 AI for Science Workshop},
  year={2021}
}

Please also cite the relevant frameworks: PyG, PyTorch Distributed, PyTorch Lightning,

and any extensions you use: 🤗 , DeepChem, NFFs, etc.

Extensions

When you're ready to upgrade to fully configurable, reproducible, and scalable workflows, use hydra-zen. hydra-zen integrates seamlessly with LitMatter to self-document ML experiments and orchestrate multiple training runs for extensive hyperparameter sweeps.

Disclaimer

DISTRIBUTION STATEMENT A. Approved for public release. Distribution is unlimited.

© 2021 MASSACHUSETTS INSTITUTE OF TECHNOLOGY

Subject to FAR 52.227-11 – Patent Rights – Ownership by the Contractor (May 2014)
SPDX-License-Identifier: MIT

This material is based upon work supported by the Under Secretary of Defense for Research and Engineering under Air Force Contract No. FA8702-15-D-0001. Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the Under Secretary of Defense for Research and Engineering.

The software/firmware is provided to you on an As-Is basis.

Owner
Nathan Frey
Postdoc at MIT
Nathan Frey
SMPL-X: A new joint 3D model of the human body, face and hands together

SMPL-X: A new joint 3D model of the human body, face and hands together [Paper Page] [Paper] [Supp. Mat.] Table of Contents License Description News I

Vassilis Choutas 1k Jan 09, 2023
An implementation of the 1. Parallel, 2. Streaming, 3. Randomized SVD using MPI4Py

PYPARSVD This implementation allows for a singular value decomposition which is: Distributed using MPI4Py Streaming - data can be shown in batches to

Romit Maulik 44 Dec 31, 2022
Official code for 'Robust Siamese Object Tracking for Unmanned Aerial Manipulator' and offical introduction to UAMT100 benchmark

SiamSA: Robust Siamese Object Tracking for Unmanned Aerial Manipulator Demo video 📹 Our video on Youtube and bilibili demonstrates the evaluation of

Intelligent Vision for Robotics in Complex Environment 12 Dec 18, 2022
This is the repository of shape matching algorithm Iterative Rotations and Assignments (IRA)

Description This is the repository of shape matching algorithm Iterative Rotations and Assignments (IRA), described in the publication [1]. Directory

MAMMASMIAS Consortium 6 Nov 14, 2022
pip install python-office

🍬 python for office 👉 http://www.python4office.cn/ 👈 🌎 English Documentation 📚 简介 Python-office 是一个 Python 自动化办公第三方库,能解决大部分自动化办公的问题。而且每个功能只需一行代码,

程序员晚枫 272 Dec 29, 2022
Official code for our ICCV paper: "From Continuity to Editability: Inverting GANs with Consecutive Images"

GANInversion_with_ConsecutiveImgs Official code for our ICCV paper: "From Continuity to Editability: Inverting GANs with Consecutive Images" https://a

QingyangXu 38 Dec 07, 2022
YOLOX-RMPOLY

本算法为适应robomaster比赛,而改动自矩形识别的yolox算法。 基于旷视科技YOLOX,实现对不规则四边形的目标检测 TODO 修改onnx推理模型 更改/添加标注: 1.yolox/models/yolox_polyhead.py: 1.1继承yolox/models/yolo_

3 Feb 25, 2022
A GOOD REPRESENTATION DETECTS NOISY LABELS

A GOOD REPRESENTATION DETECTS NOISY LABELS This code is a PyTorch implementation of the paper: Prerequisites Python 3.6.9 PyTorch 1.7.1 Torchvision 0.

<a href=[email protected]"> 64 Jan 04, 2023
a minimal terminal with python 😎😉

Meterm a terminal with python 😎 How to use Clone Project: $ git clone https://github.com/motahharm/meterm.git Run: in Terminal: meterm.exe Or pip ins

Motahhar.Mokfi 5 Jan 28, 2022
A PyTorch implementation: "LASAFT-Net-v2: Listen, Attend and Separate by Attentively aggregating Frequency Transformation"

LASAFT-Net-v2 Listen, Attend and Separate by Attentively aggregating Frequency Transformation Woosung Choi, Yeong-Seok Jeong, Jinsung Kim, Jaehwa Chun

Woosung Choi 29 Jun 04, 2022
Buffon’s needle: one of the oldest problems in geometric probability

Buffon-s-Needle Buffon’s needle is one of the oldest problems in geometric proba

3 Feb 18, 2022
JFB: Jacobian-Free Backpropagation for Implicit Models

JFB: Jacobian-Free Backpropagation for Implicit Models

Typal Research 28 Dec 11, 2022
An open-source Kazakh named entity recognition dataset (KazNERD), annotation guidelines, and baseline NER models.

Kazakh Named Entity Recognition This repository contains an open-source Kazakh named entity recognition dataset (KazNERD), named entity annotation gui

ISSAI 9 Dec 23, 2022
A supplementary code for Editable Neural Networks, an ICLR 2020 submission.

Editable neural networks A supplementary code for Editable Neural Networks, an ICLR 2020 submission by Anton Sinitsin, Vsevolod Plokhotnyuk, Dmitry Py

Anton Sinitsin 32 Nov 29, 2022
Projecting interval uncertainty through the discrete Fourier transform

Projecting interval uncertainty through the discrete Fourier transform This repo

1 Mar 02, 2022
A simple image/video to Desmos graph converter run locally

Desmos Bezier Renderer A simple image/video to Desmos graph converter run locally Sample Result Setup Install dependencies apt update apt install git

Kevin JY Cui 339 Dec 23, 2022
Companion code for the paper "An Infinite-Feature Extension for Bayesian ReLU Nets That Fixes Their Asymptotic Overconfidence" (NeurIPS 2021)

ReLU-GP Residual (RGPR) This repository contains code for reproducing the following NeurIPS 2021 paper: @inproceedings{kristiadi2021infinite, title=

Agustinus Kristiadi 4 Dec 26, 2021
Official implementation of Long-Short Transformer in PyTorch.

Long-Short Transformer (Transformer-LS) This repository hosts the code and models for the paper: Long-Short Transformer: Efficient Transformers for La

NVIDIA Corporation 198 Dec 29, 2022