Generative Modeling of High-resolution Global Precipitation Forecasts

no code implementations • 22 Oct 2022 • James Duncan, Shashank Subramanian, Peter Harrington

Forecasting global precipitation patterns and, in particular, extreme precipitation events is of critical importance to preparing for and adapting to climate change.

Weather Forecasting

FourCastNet: Accelerating Global High-Resolution Weather Forecasting using Adaptive Fourier Neural Operators

no code implementations • 8 Aug 2022 • Thorsten Kurth, Shashank Subramanian, Peter Harrington, Jaideep Pathak, Morteza Mardani, David Hall, Andrea Miele, Karthik Kashinath, Animashree Anandkumar

Extreme weather amplified by climate change is causing increasingly devastating impacts across the globe.

Weather Forecasting

FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators

1 code implementation • 22 Feb 2022 • Jaideep Pathak, Shashank Subramanian, Peter Harrington, Sanjeev Raja, Ashesh Chattopadhyay, Morteza Mardani, Thorsten Kurth, David Hall, Zongyi Li, Kamyar Azizzadenesheli, Pedram Hassanzadeh, Karthik Kashinath, Animashree Anandkumar

FourCastNet accurately forecasts high-resolution, fast-timescale variables such as the surface wind speed, precipitation, and atmospheric water vapor.

Weather Forecasting

Self-supervised similarity search for large scientific datasets

2 code implementations • 25 Oct 2021 • George Stein, Peter Harrington, Jacqueline Blaum, Tomislav Medan, Zarija Lukic

We present the use of self-supervised learning to explore and exploit large unlabeled datasets.

Self-Supervised Learning Semantic Similarity +1

Mining for Strong Gravitational Lenses with Self-supervised Learning

1 code implementation • 30 Sep 2021 • George Stein, Jacqueline Blaum, Peter Harrington, Tomislav Medan, Zarija Lukic

We employ self-supervised representation learning to distill information from 76 million galaxy images from the Dark Energy Spectroscopic Instrument Legacy Imaging Surveys' Data Release 9.

Representation Learning Self-Supervised Learning

Fast, high-fidelity Lyman $α$ forests with convolutional neural networks

1 code implementation • 23 Jun 2021 • Peter Harrington, Mustafa Mustafa, Max Dornfest, Benjamin Horowitz, Zarija Lukić

Full-physics cosmological simulations are powerful tools for studying the formation and evolution of structure in the universe but require extreme computational resources.

Estimating Galactic Distances From Images Using Self-supervised Representation Learning

no code implementations • 12 Jan 2021 • Md Abul Hayat, Peter Harrington, George Stein, Zarija Lukić, Mustafa Mustafa

We use a contrastive self-supervised learning framework to estimate distances to galaxies from their photometric images.

Representation Learning Self-Supervised Learning

Self-Supervised Representation Learning for Astronomical Images

1 code implementation • 24 Dec 2020 • Md Abul Hayat, George Stein, Peter Harrington, Zarija Lukić, Mustafa Mustafa

We show that, without the need for labels, self-supervised learning recovers representations of sky survey images that are semantically useful for a variety of scientific tasks.

Astronomy Contrastive Learning +4

The Case for Strong Scaling in Deep Learning: Training Large 3D CNNs with Hybrid Parallelism

1 code implementation • 25 Jul 2020 • Yosuke Oyama, Naoya Maruyama, Nikoli Dryden, Erin McCarthy, Peter Harrington, Jan Balewski, Satoshi Matsuoka, Peter Nugent, Brian Van Essen

We present scalable hybrid-parallel algorithms for training large-scale 3D convolutional neural networks.