uqdeepnn 0.1.17

PyTorch uncertainty quantification toolkit with deep ensembles, Bayes-by-Backprop VI, Laplace, SGLD, MC Dropout, Gaussian Processes, and scientific ML backbones.

Deep-UQ is a PyTorch toolkit for uncertainty-aware machine learning.

It collects practical uncertainty quantification methods, Gaussian-process models, and scientific machine learning backbones in one package, with a common focus on predictive uncertainty for regression, classification, and field-to-field surrogate modeling.

Install

pip install uqdeepnn

Import

import deepuq

Included methods

• Deep Ensembles
• Variational Inference (Bayes by Backprop, heteroscedastic VI, multi-output VI, and last-layer VI)
• Laplace Approximation
• MCMC via SGLD
• MC Dropout
• Gaussian Processes

Scientific machine learning backbones

• DeepONet
• Fourier Neural Operator (FNO)
• CNN / ResNet spatial surrogates
• U-Net backbones
• Physics-Informed Neural Networks (PINNs)

Documentation

• Docs: https://vispikarkaria.github.io/Deep-UQ/
• Tutorials: https://vispikarkaria.github.io/Deep-UQ/tutorials/
• API reference: https://vispikarkaria.github.io/Deep-UQ/api/
• Variational Inference guide: https://vispikarkaria.github.io/Deep-UQ/methods/variational-inference/
• GitHub: https://github.com/Vispikarkaria/Deep-UQ

Package names

• PyPI package: uqdeepnn
• Python import: deepuq
• Project / docs name: Deep-UQ