swMATH ID: 36444
Software Authors: Krueger, David; Maharaj, Tegan; Kramár, János; Pezeshki, Mohammad; Ballas, Nicolas; Ke, Nan Rosemary; Goyal, Anirudh; Bengio, Yoshua; Courville, Aaron; Pal, Chris
Description: Zoneout: Regularizing RNNs by Randomly Preserving Hidden Activations. We propose zoneout, a novel method for regularizing RNNs. At each timestep, zoneout stochastically forces some hidden units to maintain their previous values. Like dropout, zoneout uses random noise to train a pseudo-ensemble, improving generalization. But by preserving instead of dropping hidden units, gradient information and state information are more readily propagated through time, as in feedforward stochastic depth networks. We perform an empirical investigation of various RNN regularizers, and find that zoneout gives significant performance improvements across tasks. We achieve competitive results with relatively simple models in character- and word-level language modelling on the Penn Treebank and Text8 datasets, and combining with recurrent batch normalization yields state-of-the-art results on permuted sequential MNIST.
Homepage: https://arxiv.org/abs/1606.01305
Source Code:  https://github.com/teganmaharaj/zoneout
Dependencies: Mathematica
Keywords: arXiv_cs.NE; arXiv_cs.CL; Machine Learning; arXiv_cs.LG; Regularizing RNNs; Preserving Hidden Activations; zoneout
Related Software: Adam; GloVe; Clockwork RNN; ImageNet; MNIST; astsa; LSTM; IMDB; UCF101; word2vec; Skip RNN; BinaryConnect; Theano; RMSprop; Penn Treebank; SGDR; GELUs; Drop-Activation; mixup; AutoAugment
Cited in: 3 Documents

Standard Articles

1 Publication describing the Software Year
Zoneout: Regularizing RNNs by Randomly Preserving Hidden Activations arXiv
David Krueger, Tegan Maharaj, János Kramár, Mohammad Pezeshki, Nicolas Ballas, Nan Rosemary Ke, Anirudh Goyal, Yoshua Bengio, Aaron Courville, Chris Pal

Citations by Year