API Reference

Reference documentation for the pinnrl package, generated from the source docstrings with mkdocstrings.

Module	Description
Configuration	Dataclasses built from config.yaml (Config, TrainingConfig, ModelConfig, …)
Neural Networks	The PINNModel factory and the 7 architecture backends
PDEs	The PDE registry, PDEBase and the 9 PDE implementations
Training	PDETrainer (train, train_steps), adaptive loss weights, build_model, load_run
RL sampler	The experimental residual-bandit collocation sampler and its DQN
Plots	Solution, training-history and collocation plots
Benchmarks	Numerical reference solutions and the equal-compute sampling study