AnyPINN

Solve differential equations with Physics-Informed Neural Networks.

Most PINN libraries make you wire up every loss term, collocation grid, and training loop by hand before you see a single result. AnyPINN gives you a working experiment in one command and then lets you peel back every layer when you're ready.

One Command Setup

anypinn create scaffolds a runnable project from 16 built-in templates. No boilerplate, no wiring. Edit physics and press start.

Get started
Inverse-First

Recover unknown parameters from observations. Promoting a constant to a learnable parameter is a one-line change.

How it works
Training Agnostic

Problem is a plain nn.Module. Use PyTorch Lightning, a raw training loop, or anything that calls .backward().

Lightning vs Core

Quick Start

uvx anypinn create my-project

anypinn create scaffolds a complete, runnable project with your choice of template, data source, and training framework.

Full walkthrough

Who Is This For?

Experimenter

Run a known problem, tweak parameters, see results. Pick a built-in template, change config.py, press start.
Researcher

Define new physics or custom constraints. Subclass Constraint and Problem, use the provided loss machinery.
Framework Builder

Custom training loops, novel architectures. Use anypinn.core directly, no Lightning required.