We introduce two Python frameworks to train neural networks on large datasets: Blocks and Fuel. Blocks is based on Theano, a linear algebra compiler with CUDA-support. It facilitates the training of complex neural network models by providing parametrized Theano operations, attaching metadata to Theano's symbolic computational graph, and providing an extensive set of utilities to assist training the networks, e.g. training algorithms, logging, monitoring, visualization, and serialization. Fuel provides a standard format for machine learning datasets. It allows the user to easily iterate over large datasets, performing many types of pre-processing on the fly.
@article{arxiv.1506.00619,
title = {Blocks and Fuel: Frameworks for deep learning},
author = {Bart van Merriënboer and Dzmitry Bahdanau and Vincent Dumoulin and Dmitriy Serdyuk and David Warde-Farley and Jan Chorowski and Yoshua Bengio},
journal= {arXiv preprint arXiv:1506.00619},
year = {2015}
}