Giter Site home page Giter Site logo

jxzhangjhu / implicit-normalizing-flows Goto Github PK

View Code? Open in Web Editor NEW

This project forked from thu-ml/implicit-normalizing-flows

0.0 0.0 0.0 237 KB

Code for "Implicit Normalizing Flows" (ICLR 2021 spotlight)

License: MIT License

Python 99.63% Shell 0.37%

implicit-normalizing-flows's Introduction

Implicit Normalizing Flows (ICLR 2021 Spotlight)[arxiv][slides]

This repository contains Pytorch implementation of experiments from the paper Implicit Normalizing Flows. The implementation is based on Residual Flows.

Implicit Normalizing Flows generalize normalizing flows by allowing the invertible mapping to be implicitly defined by the roots of an equation F(z, x) = 0. Building on Residual Flows, we propose:

  • A unique and invertible mapping defined by an equation of the latent variable and the observed variable.
  • A more powerful function space than Residual Flows, which relaxing the Lipschitz constraints of Residual Flows.
  • A scalable algorithm for generative modeling.

As an example, in 1-D case, the function space of ImpFlows contains all strongly monotonic differentiable functions, i.e.

A 1-D function fitting example:

Requirements

  • PyTorch 1.4, torchvision 0.5
  • Python 3.6+

Density Estimation Experiments

NOTE: By default, O(1)-memory gradients are enabled. However, the logged bits/dim during training will not be an actual estimate of bits/dim but whatever scalar was used to generate the unbiased gradients. If you want to check the actual bits/dim for training (and have sufficient GPU memory), set --neumann-grad=False. Note however that the memory cost can stochastically vary during training if this flag is False.

Toy 2-D data:

bash run_toy.sh

CIFAR10:

bash run_cifar10.sh

Tabular Data:

bash run_tabular.sh

BibTeX

To be done after ICLR 2021.

implicit-normalizing-flows's People

Contributors

luchengthu avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.