bayesiains/nflows
A PyTorch library providing comprehensive implementations of normalizing flow models for density estimation and generative tasks.
Not currently ranked — collecting fresh signals.
star history
nflows is a collection of normalizing flow implementations built on PyTorch. It provides invertible transformations, base distributions, and flow compositions that enable users to define and train normalizing flow models. The library supports computing log probabilities of inputs and sampling from the learned distribution, making it suitable for generative modeling and density estimation tasks.
Frequently asked
- What is bayesiains/nflows?
- A PyTorch library providing comprehensive implementations of normalizing flow models for density estimation and generative tasks.
- Is nflows open source?
- Yes — bayesiains/nflows is open source, released under the MIT license.
- What language is nflows written in?
- bayesiains/nflows is primarily written in Python.
- How popular is nflows?
- bayesiains/nflows has 1k stars on GitHub.
- Where can I find nflows?
- bayesiains/nflows is on GitHub at https://github.com/bayesiains/nflows.