executor¶

The Executor.

An object that, given an ashpy.contexts.BaseContext, carries a function and the way of executing it.

Classes

`Executor`	Carry a function and the way of executing it.
`SumExecutor`	The sum executor.

class ashpy.losses.executor.Executor(fn=None)[source]¶

Carry a function and the way of executing it. Given a context.

__init__(fn=None)[source]¶

Initialize the Executor.

abstract call(context, **kwargs)[source]¶

Execute the function, using the information provided by the context.

Parameters: context (ashpy.contexts.BaseContext) – The function execution Context.
Return type: Tensor
Returns: tf.Tensor – Output Tensor.

property fn¶

Return the Keras loss function to execute.

property global_batch_size¶

Global batch size comprises the batch size for each cpu.

Calculated as batch_size_for_replica*replica_numbers.

static reduce_loss(call_fn)[source]¶

Create a Decorator to reduce Losses. Used to simplify things.

Apply a reduce sum operation to the loss and divide the result by the batch size.

property weight¶

Return the loss weight.

This weight is multiplied by the loss value. This is useful when working with multiples losses.

class ashpy.losses.executor.SumExecutor(executors)[source]¶

The sum executor. Executes the call of each fn and weights the losses.

Each Executor gets called (thus reducing its carried function), the results are then summed together.

__init__(executors)[source]¶

Initialize the SumExecutor.

Parameters: executors (list of [ashpy.executors.Executor]) – Array of ashpy.executors.Executor to sum evaluate and sum together.
Return type: None
Returns: None

call(*args, **kwargs)[source]¶

Evaluate and sum together the Executors.

property executors¶

Return the List of Executors.

property global_batch_size¶

Global batch size comprises the batch size for each cpu.

Calculated as batch_size_for_replica*replica_numbers.