pyro: Deep universal probabilistic programming with Python and PyTorch

Discussion topics for individual codes
Post Reply
Ada Coda
ASCL Robot
Posts: 2161
Joined: Thu May 08, 2014 5:37 am

pyro: Deep universal probabilistic programming with Python and PyTorch

Post by Ada Coda » Mon Nov 01, 2021 2:43 am

pyro: Deep universal probabilistic programming with Python and PyTorch

Abstract: Pyro is a flexible, scalable deep probabilistic programming library built on PyTorch. It can represent any computable probability distribution and scales to large data sets with little overhead compared to hand-written code. The library is implemented with a small core of powerful, composable abstractions. Its high-level abstractions express generative and inference models, but also allows experts to customize inference.

Credit: Bingham, Eli; Chen, Jonathan P.; Jankowiak, Martin; Obermeyer, Fritz; Pradhan, Neeraj; Karaletsos, Theofanis; Singh, Rohit; Szerlip, Paul; Horsfall, Paul; Goodman, Noah D.

Site: https://github.com/pyro-ppl/pyro
https://ui.adsabs.harvard.edu/abs/2020MNRAS.496..381C

Bibcode: 2021ascl.soft10016B

Preferred citation method: https://jmlr.org/papers/v20/18-403.html

ID: ascl:2110.016

Post Reply