weka-lasagne

WekaLasagne is a wrapper for the neural network library Lasagne, which allows users to easily construct neural networks using WEKA's GUI.

Examples

This repository comes with the MNIST digits dataset in ARFF format. We can easily define a multilayer perceptron and train it, using 66% of the data for training and the rest for testing, like so:

java weka.Run .LasagneNet \
     -layer ".DenseLayer -num_units 50 -nonlinearity .Rectify" \
     -loss ".CategoricalCrossEntropy" \
     -update ".Momentum -learning_rate 0.01 -momentum 0.9" -epochs 100 -bs 128 \
     -out_file "/tmp/output.txt" \
     -t datasets/mnist.arff -split-percentage 66