tasks

A procedure that consumes and produces well-defined components is termed a MAITE task. All below tasks are valid across AI problems and will use static type checking to validate that all passed components are structurally compatible.

tasks.evaluate(*, model[, metric, ...])	Evaluate a model's performance on data according to some metric with optional augmentation.
tasks.predict(*, model[, dataloader, ...])	Make predictions for a given model & data source with optional augmentation.
tasks.evaluate_from_predictions(*, metric, ...)	Evaluate pre-calculated predictions against target (truth) data for some specified metric.
tasks.augment_dataloader(*, augmentation, ...)	Create an DataLoader of augmented inputs from a Dataset or DataLoader.