API reference

Generated from the docstrings of the peal package. Only first-party code is listed; the vendored research code under peal/dependencies/ keeps its own documentation.

peal.adaptors	Adaptors: the procedures that repair a predictor from explanation feedback.
peal.architectures	Predictor architectures and the building blocks they are assembled from.
peal.data	Datasets, dataloaders and the DataConfig schema they are built from.
peal.editors	Editors: latent-inversion recipes applied to an existing image.
peal.entrypoints	Command-line entry points of PEAL.
peal.explainers	Explainers: methods that explain a predictor's decisions.
peal.generators	Generative models that render the counterfactuals the explainers need.
peal.sparse_dictionaries	Sparse dictionaries: concept bases over an encoder's activation space.
peal.teachers	Teachers: the sources of feedback an adaptor learns from.
peal.training	Training of predictors: trainers, criterions, loggers and their configs.
peal.visualization	Rendering helpers the pipeline uses, plus the DiDAE table-1 script.
peal.web	PEAL web demo: upload an ONNX classifier and a zipped image-folder dataset, get the DiDAE Clever Hans analysis and a corrected classifier back.
peal.global_utils	Cross-cutting helpers shared by the whole PEAL library.
peal.preflight	Check whether a run config could run, without starting it.