tfm.core.actions.RecoveryAction

Stay organized with collections Save and categorize content based on your preferences.

Train action to recover from loss blowup.

tfm.core.actions.RecoveryAction(
    checkpoint_manager: tf.train.CheckpointManager
)

Checks the loss value by the given threshold. If applicable, recover the model by reading the checkpoint on disk.

Methods

__call__

View source

__call__(
    _
)

Recovers the training by triggering checkpoint restoration.