Resource retrieval
Retrieve the pre-trained net:
Out[1]= |  |
Basic usage
Apply the trained net to a set of inputs:
Out[2]= |  |
Give class probabilities for a single input:
Out[3]= |  |
Feature extraction
Create a subset of the MNIST dataset:
Out[4]= |  |
Remove the last linear layer of the net, which will be used as a feature extractor:
Out[5]= |  |
Visualize the features of a subset of the MNIST dataset:
Out[6]= |  |
Visualization of net operation
Extract the convolutional features from the first layer:
Out[7]= |  |
Visualize the features:
Out[8]= |  |
Training the uninitialized architecture
Retrieve the uninitialized architecture:
Out[9]= |  |
Retrieve the MNIST dataset:
Out[10]= |  |
Use the training dataset provided:
Out[11]= |  |
Use the test dataset provided:
Out[12]= |  |
Train the net:
Out[13]= |  |
Generate a ClassifierMeasurementsObject of the net with the test set:
Out[14]= |  |
Evaluate the accuracy on the validation set:
Out[15]= |  |
Visualize the confusion matrix:
Out[16]= |  |
Net information
Inspect the number of parameters of all arrays in the net:
Out[17]= |  |
Obtain the total number of parameters:
Out[18]= |  |
Obtain the layer type counts:
Out[19]= |  |
Display the summary graphic:
Out[20]= |  |
Export to MXNet
Export the net into a format that can be opened in MXNet:
Out[21]= |  |
Export also creates a net.params file containing parameters:
Out[22]= |  |
Get the size of the parameter file:
Out[23]= |  |
The size is similar to the byte count of the resource object:
Out[24]= |  |
Represent the MXNet net as a graph:
Out[25]= |  |