TrainingCallback

class TrainingCallback[source]

Bases: object

An interface for training callbacks.

Initialize the callback.

Attributes Summary

`loss`	The loss, accessed via the training loop.
`model`	The model, accessed via the training loop.
`optimizer`	The optimizer, accessed via the training loop.
`result_tracker`	The result tracker, accessed via the training loop.
`training_loop`	The training loop.

Methods Summary

`on_batch`(epoch, batch, batch_loss, **kwargs)	Call for training batches.
`post_batch`(epoch, batch, **kwargs)	Call for training batches.
`post_epoch`(epoch, epoch_loss, **kwargs)	Call after epoch.
`post_train`(losses, **kwargs)	Call after training.
`pre_step`(**kwargs)	Call before the optimizer's step.
`register_training_loop`(training_loop)	Register the training loop.

Attributes Documentation

loss

The loss, accessed via the training loop.

Return type: Loss

model

The model, accessed via the training loop.

Return type: Model

optimizer

The optimizer, accessed via the training loop.

Return type: Optimizer

result_tracker

The result tracker, accessed via the training loop.

Return type: ResultTracker

training_loop: The training loop.

Methods Documentation

on_batch(epoch, batch, batch_loss, **kwargs)[source]

Call for training batches.

Return type

None

Parameters

epoch (int) –
batch_loss (float) –
kwargs (Any) –

post_batch(epoch, batch, **kwargs)[source]

Call for training batches.

Return type

None

Parameters

epoch (int) –
kwargs (Any) –

post_epoch(epoch, epoch_loss, **kwargs)[source]

Call after epoch.

Return type

None

Parameters

epoch (int) –
epoch_loss (float) –
kwargs (Any) –

post_train(losses, **kwargs)[source]

Call after training.

Return type

None

Parameters

losses (List[float]) –
kwargs (Any) –

pre_step(**kwargs)[source]

Call before the optimizer’s step.

Return type: None
Parameters: kwargs (Any) –

register_training_loop(training_loop)[source]

Register the training loop.

Return type: None