controllers

This is summary documentation for hippynn.experiment.controllers module. Click here for full documentation.

Links to Full Documentation:

Managing optimizer, scheduler, and end of training

Functions

accepts_metrics(sch)	Return if a scheduler accepts the validation metric.
is_scheduler_like(thing)	Test if an object can be used as a scheduler.

Classes

Controller(optimizer, scheduler, batch_size, ...)	Class for controlling the training dynamics.
PatienceController(*args, ...)	A subclass of Controller that terminates if training has not improved for a given number of epochs.
RaiseBatchSizeOnPlateau(optimizer, ...[, ...])	Learning rate scheduler compatible with pytorch schedulers.