Blog

Apache MXNet’s module API is like a FeedForward model and it is easier to compose similar to Torch module. It consists of following classes −

BaseModule([logger])

It represents the base class of a module. A module can be thought of as computation component or computation machine. The job of a module is to execute forward and backward passes. It also updates parameters in a model.

Methods

Following table shows the methods consisted in BaseModule class−This method will get states from all devices

Attributes

Following table shows the attributes consisted in the methods of BaseModule class −

data_shapes: You can refer the link available at https://mxnet.apache.org for details. output_shapes: More

output_shapes: More information is available at https://mxnet.apache.org/api/python

BucketingModule(sym_gen[…])

It represents the Bucketingmodule class of a Module which helps to deal efficiently with varying length inputs.

Methods

Following table shows the methods consisted in BucketingModule class −

Attributes

Following table shows the attributes consisted in the methods of BaseModule class −

data_shapes − You can refer the link at https://mxnet.apache.org/api/python/docs for more information.

output_shapes− You can refer the link at https://mxnet.apache.org/api/python/docs for more information.

Module(symbol[,data_names, label_names,…])

It represents a basic module that wrap a symbol.

Methods

Following table shows the methods consisted in Module class −

Attributes

Following table shows the attributes consisted in the methods of Module class −

data_shapes: Visit the link https://mxnet.apache.org/api/python/docs/api/module for further details.

output_shapes: The link given herewith https://mxnet.apache.org/api/python/docs/api/module/index.html will offer other important information.

PythonLossModule([name,data_names,…])

The base of this class is mxnet.module.python_module.PythonModule. PythonLossModule class is a convenient module class which implements all or many of the module APIs as empty functions.

Methods

Following table shows the methods consisted in PythonLossModule class:

PythonModule([data_names,label_names…])

The base of this class is mxnet.module.base_module.BaseModule. PythonModule class also is a convenient module class which implements all or many of the module APIs as empty functions.

Methods

Following table shows the methods consisted in PythonModule class −

Attributes

Following table shows the attributes consisted in the methods of PythonModule class −

data_shapes − Follow the link https://mxnet.apache.org for details.

output_shapes − For more details, visit the link available at https://mxnet.apache.org

SequentialModule([logger])

The base of this class is mxnet.module.base_module.BaseModule. SequentialModule class also is a container module that can chain more than two (multiple) modules together.

Methods

Following table shows the methods consisted in SequentialModule class

Attributes

Following table shows the attributes consisted in the methods of BaseModule class −

data_shapes − The link given herewith https://mxnet.apache.org will help you in understanding the attribute in much detail.

output_shapes − Follow the link available at https://mxnet.apache.org/api for details.

Implementation Examples

In the example below, we are going create a mxnet module.

import mxnet as mx
input_data = mx.symbol.Variable('input_data')
f_connected1 = mx.symbol.FullyConnected(data, name='f_connected1', num_hidden=128)
activation_1 = mx.symbol.Activation(f_connected1, name='relu1', act_type="relu")
f_connected2 = mx.symbol.FullyConnected(activation_1, name = 'f_connected2', num_hidden = 64)
activation_2 = mx.symbol.Activation(f_connected2, name='relu2',
act_type="relu")
f_connected3 = mx.symbol.FullyConnected(activation_2, name='fc3', num_hidden=10)
out = mx.symbol.SoftmaxOutput(f_connected3, name = 'softmax')
mod = mx.mod.Module(out)
print(out)

Output

The output is mentioned below −

<Symbol softmax>

Example

print(mod)

Output

The output is shown below −

<mxnet.module.module.Module object at 0x00000123A9892F28>

In this example below, we will be implementing forward computation

import mxnet as mx
from collections import namedtuple
Batch = namedtuple('Batch', ['data'])
data = mx.sym.Variable('data')
out = data * 2
mod = mx.mod.Module(symbol=out, label_names=None)
mod.bind(data_shapes=[('data', (1, 10))])
mod.init_params()
data1 = [mx.nd.ones((1, 10))]
mod.forward(Batch(data1))
print (mod.get_outputs()[0].asnumpy())

Output

When you execute the above code, you should see the following output −

[[2. 2. 2. 2. 2. 2. 2. 2. 2. 2.]]

Example

data2 = [mx.nd.ones((3, 5))]

mod.forward(Batch(data2))
print (mod.get_outputs()[0].asnumpy())

Output

Given below is the output of the code −

[[2. 2. 2. 2. 2.]
[2. 2. 2. 2. 2.]
[2. 2. 2. 2. 2.]]