Callback¶

Inheritance Diagram

class ashpy.callbacks.callback.Callback(name)[source]¶

Bases: tensorflow.python.module.module.Module

Callback definition.

Every callback must extend from this class. This class defines the basic events. Every event takes as input the context in order to use the objects defined. Inheritance from tf.Module is required since callbacks have a state

Order: .. code-block:

--on_train_start

----on_epoch_start

------on_batch_start

------on_batch_end

----on_epoch_end

--on_train_end

on_exception – if an Exception was raised

on_event - Called when an event is triggered

Methods

`__init__`(name)	Initialize the Callback.
`on_batch_end`(context)	Handle on_batch_end event.
`on_batch_start`(context)	Handle on_batch_start event.
`on_epoch_end`(context)	Handle on_epoch_end event.
`on_epoch_start`(context)	Handle on_epoch_start event.
`on_event`(event, context)	Handle the on_event event.
`on_exception`(context)	Handle on_exception event.
`on_train_end`(context)	Handle on_train_end event.
`on_train_start`(context)	Handle on_train_start event.

Attributes

`name`	Return the name of the callback.
`name_scope`	Returns a tf.name_scope instance for this class.
`submodules`	Sequence of all sub-modules.
`trainable_variables`	Sequence of variables owned by this module and it’s submodules.
`variables`	Sequence of variables owned by this module and it’s submodules.

__init__(name)[source]¶

Initialize the Callback.

Parameters:	name (str) – Callback name.

Warning

When using multiple callbacks with the same trainer make sure they have different ids.

Return type:	`None`

name¶: Return the name of the callback.

on_batch_end(context)[source]¶

Handle on_batch_end event.

Method called at the end of a batch.

Parameters:	context (`ashpy.contexts.context.Context`) – training context
Return type:	`None`

on_batch_start(context)[source]¶

Handle on_batch_start event.

Method called at the beginning of a batch.

Parameters:	context (`ashpy.contexts.context.Context`) – training context
Return type:	`None`

on_epoch_end(context)[source]¶

Handle on_epoch_end event.

Method called at the end of an epoch.

Parameters:	context (`ashpy.contexts.context.Context`) – training context
Return type:	`None`

on_epoch_start(context)[source]¶

Handle on_epoch_start event.

Method called at the beginning of an epoch.

Parameters:	context (`ashpy.contexts.context.Context`) – training context
Return type:	`None`

on_event(event, context)[source]¶

Handle the on_event event.

Method called when an event is triggered.

Parameters:	event (`ashpy.callbacks.events.Event`) – triggered event context (`ashpy.contexts.context.Context`) – training context
Return type:	`None`

on_exception(context)[source]¶

Handle on_exception event.

Method called when an exception is raised.

Parameters:	context (`ashpy.contexts.context.Context`) – training context
Return type:	`None`

on_train_end(context)[source]¶

Handle on_train_end event.

Method called at the end of the training loop.

Parameters:	context (`ashpy.contexts.context.Context`) – training context
Return type:	`None`

on_train_start(context)[source]¶

Handle on_train_start event.

Method called at the beginning of the training loop.

Parameters:	context (`ashpy.contexts.context.Context`) – training context
Return type:	`None`