openai/glow
OfficialCode for reproducing results in "Glow: Generative Flow with Invertible 1x1 Convolutions"
Our Review
- •Reproduces SOTA generative flow results
- •Supports large datasets and distributed training
- •Interactive demo for manipulations
Our Verdict
Reproduction code for Glow: generative flow model with invertible 1x1 convolutions. For ML researchers in density estimation. Key advantage: exact likelihoods over GANs.
Frequently Asked Questions
What is openai/glow used for?
It provides Python code to reproduce results from the Glow paper on generative flows using invertible 1x1 convolutions for tasks like image density estimation and generation.
What is glow?
Code for reproducing results in "Glow: Generative Flow with Invertible 1x1 Convolutions"
How do I install glow?
Visit the GitHub repository at https://github.com/openai/glow for installation instructions.
What license does glow use?
glow uses the MIT license.
What are alternatives to glow?
Search My AI Guide for similar tools in this category.
Open source & community-verified
MIT licensed: free to use in any project, no strings attached. 3,184 developers have starred this, meaning the community has reviewed and trusted it.
Reviewed by My AI Guide for relevance, quality, and active maintenance before listing.