This repository was archived by the owner on Jul 10, 2021. It is now read-only.

Description
The tool slowly consumes all available memory (24GB) during fit(). Version 0.7.
Using as:
model = Classifier(
layers=[
Layer("Rectifier", units=256),
Layer("Rectifier", units=256),
Layer("Rectifier", units=256),
Layer("Softmax")
],
learning_rule='adagrad',
batch_size=32,
n_iter=100,
regularize="L1",
weight_decay=0.001,
verbose=True
)