Train Neural Networks

TODO

Train a neural network.

nntoolkit.train.get_parser()

Return the parser object for this script.

nntoolkit.train.main(model_file, model_output_file, training_data, batch_size, learning_rate, epochs)

Train model_file with training_data.

nntoolkit.train.minibatch_gradient_descent(model, x, y, batch_size=256, lr=0.1, epochs=100)

Train a given neural network.

Parameters:

• model – A parsed model
• x – A numpy array with features
• y – The list of classes
• batch_size – A positive integer which defines after how many training examples the values of the neural network get adjusted
• lr – Learning rate. Has to be positive.