CallbackEvents

class hybrid_learning.concepts.train_eval.callbacks.CallbackEvents(value)

Bases: Enum

Supported callback events for training and evaluation. Events can be:

• before_epoch_train: right before iterating one epoch of the

  training set

• after_batch_train, after_epoch_eval:

  after every batch of an epoch of training

• after_epoch_train: right after iterating one epoch of the

  training set; provides kpis of the training set of current epoch, and the train_loader

• after_epoch_eval: right after iterating one epoch of evaluation

  (validation or test set); provides evaluation kpis of the current epoch and the val_loader

• after_epoch: after an epoch of training: provides kpis of

  current epoch

• between_epochs: after an epoch of a training run with several

  epochs: provides epoch context (total KPI history, current and total epoch number)

• after_train: after the training of several epochs;

  provides full training and validation history as well as total epoch number

• after_second_stage_train: like after_train only after the

  second stage training

Public Data Attributes:

BEFORE_EPOCH_TRAIN
AFTER_BATCH_TRAIN
AFTER_EPOCH_TRAIN
AFTER_BATCH_EVAL
AFTER_EPOCH_EVAL
AFTER_EPOCH
BETWEEN_EPOCHS
AFTER_TRAIN
AFTER_SECOND_STAGE_TRAIN

Inherited from : py: class:Enum

Special Methods:

__new__(value)

Inherited from : py: class:Enum

__repr__()	Return repr(self).
__str__()	Return str(self).
__dir__()	Returns all members and all public methods
__format__(format_spec)	Returns format using actual value type unless __str__ has been overridden.
__hash__()	Return hash(self).
__reduce_ex__(proto)	Helper for pickle.

---

AFTER_BATCH_EVAL = 'after_batch_eval'

AFTER_BATCH_TRAIN = 'after_batch_train'

AFTER_EPOCH = 'after_epoch'

AFTER_EPOCH_EVAL = 'after_epoch_eval'

AFTER_EPOCH_TRAIN = 'after_epoch_train'

AFTER_SECOND_STAGE_TRAIN = 'after_second_stage_train'

AFTER_TRAIN = 'after_train'

BEFORE_EPOCH_TRAIN = 'before_epoch_train'

BETWEEN_EPOCHS = 'between_epochs'