Trainer API

Parameters

paddle.v2.parameters.create(layers)

Create parameter pool by topology.

参数:layers – 返回:

class paddle.v2.parameters.Parameters

Parameters is a dictionary contains Paddle’s parameter. The key of Parameters is the name of parameter. The value of Parameters is a plain numpy.ndarry .

Basically usage is

data = paddle.layers.data(...)
...
out = paddle.layers.fc(...)

parameters = paddle.parameters.create(out)

parameter_names = parameters.names()
fc_mat = parameters.get('fc')
print fc_mat

keys()

keys are the names of each parameter.

返回:list of parameter name 返回类型:list

names()

names of each parameter.

返回:list of parameter name 返回类型:list

has_key(key)

has_key return true if there are such parameter name == key

参数:key (basestring) – Parameter name 返回:True if contains such key

get_shape(key)

get shape of the parameter.

参数:key (basestring) – parameter name 返回:parameter’s shape 返回类型:tuple

get(parameter_name)

Get parameter by parameter name.

Note:It will always copy the parameter from C++ side. 参数:parameter_name (basestring) – parameter name 返回:The parameter matrix. 返回类型:np.ndarray

set(parameter_name, value)

Set parameter by parameter name & matrix.

参数:

• parameter_name (basestring) – parameter name
• value (np.ndarray) – parameter matrix

返回:

Nothing.

append_gradient_machine(gradient_machine)

append gradient machine to parameters. This method is used internally in Trainer.train.

参数:gradient_machine (api.GradientMachine) – Paddle C++ GradientMachine object. 返回:

Trainer

class paddle.v2.trainer.SGD(cost, parameters, update_equation)

Simple SGD Trainer. TODO(yuyang18): Complete comments

参数:

• update_equation (paddle.v2.optimizer.Optimizer) – The optimizer object.
• cost (paddle.v2.config_base.Layer) – Target cost that neural network should be optimized.
• parameters (paddle.v2.parameters.Parameters) – The parameters dictionary.

train(reader, num_passes=1, event_handler=None, reader_dict=None)

Training method. Will train num_passes of input data.

参数:

• reader –
• num_passes – The total train passes.
• event_handler ((BaseEvent) => None) – Event handler. A method will be invoked when event occurred.

返回:

Event

All training events.

There are:

• BeginIteration
• EndIteration
• BeginPass
• EndPass

TODO(yuyang18): Complete it!

class paddle.v2.event.TestResult(evaluator, cost)

Result that trainer.test return.

class paddle.v2.event.BeginPass(pass_id)

Event On One Pass Training Start.

class paddle.v2.event.EndPass(pass_id, evaluator)

Event On One Pass Training Complete.

class paddle.v2.event.BeginIteration(pass_id, batch_id)

Event On One Batch Training Start.

class paddle.v2.event.EndIteration(pass_id, batch_id, cost, evaluator)

Event On One Batch Training Complete.