Mastering TensorFlow 1.x

Creating the TFLearn Layers

Let us learn how to create the layers of the neural network models in TFLearn:

1. Create an input layer first:

input_layer = tflearn.input_data(shape=[None,num_inputs]

2. Pass the input object to create further layers:

layer1 = tflearn.fully_connected(input_layer,10,                                 activation='relu')layer2 = tflearn.fully_connected(layer1,10,                                 activation='relu')

3. Add the output layer:

output = tflearn.fully_connected(layer2,n_classes,                                 activation='softmax')

4. Create the final net from the estimator layer such as regression:

net = tflearn.regression(output,                         optimizer='adam',                         metric=tflearn.metrics.Accuracy(),                         loss='categorical_crossentropy'                        )

The TFLearn provides several classes for layers that are described in following sub-sections. ...