Professor Forcing: A New Algorithm for Training Recurrent Networks

TL;DR

Professor Forcing introduces an adversarial training method for recurrent networks that aligns training and sampling dynamics, improving sample quality and model regularization across various sequence modeling tasks.

Contribution

It presents a novel adversarial algorithm, Professor Forcing, to better match training and sampling behaviors in recurrent networks, enhancing their generative performance.

Findings

Improves test likelihood on language and image datasets

Produces more realistic and consistent long-term samples

Acts as a regularizer to prevent overfitting

Abstract

The Teacher Forcing algorithm trains recurrent networks by supplying observed sequence values as inputs during training and using the network's own one-step-ahead predictions to do multi-step sampling. We introduce the Professor Forcing algorithm, which uses adversarial domain adaptation to encourage the dynamics of the recurrent network to be the same when training the network and when sampling from the network over multiple time steps. We apply Professor Forcing to language modeling, vocal synthesis on raw waveforms, handwriting generation, and image generation. Empirically we find that Professor Forcing acts as a regularizer, improving test likelihood on character level Penn Treebank and sequential MNIST. We also find that the model qualitatively improves samples, especially when sampling for a large number of time steps. This is supported by human evaluation of sample quality. Trade-offs between Professor Forcing and Scheduled Sampling are discussed. We produce T-SNEs showing that Professor Forcing successfully makes the dynamics of the network during training and sampling more similar.