Callbacks

Eztorch supports callbacks for trainer. It can takes callbacks from pytorch lightning or custom ones.

Custom callbacks

class eztorch.callbacks.OnlineEvaluator(optimizer, classifier, input_name='h', precision=32)

Attaches a classifier to evaluate a specific representation from the model during training.

Parameters:

• optimizer (

  Default:) – Config to instantiate an optimizer and optionally a scheduler.

• classifier (

  Default:) – Config to instantiate a classifier.

• input_name (

  Default:, optional) – Name of the representation to evaluate from the model outputs.

  Default: 'h'

• precision (

  Default:, optional) – Precision for the classifier that must match the model, if \(16\) use automatic mixed precision.

  Default: 32

Example:

optimizer = {...} # config to build an optimizer
classifier = {...} # config to build a classifier
trainer = Trainer(callbacks=[OnlineEvaluator(optimizer, classifier)])