Professor Forcing: A New Algorithm for Training Recurrent Networks

by   Alex Lamb, et al.

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.


Character-level Recurrent Neural Networks in Practice: Comparing Training and Sampling Schemes

Recurrent neural networks are nowadays successfully used in an abundance...

Scheduled Sampling for Transformers

Scheduled sampling is a technique for avoiding one of the known problems...

Auto-regressive Image Synthesis with Integrated Quantization

Deep generative models have achieved conspicuous progress in realistic i...

Tensor Networks for Language Modeling

The tensor network formalism has enjoyed over two decades of success in ...

Efficient Neural Audio Synthesis

Sequential models achieve state-of-the-art results in audio, visual and ...

Superposition of many models into one

We present a method for storing multiple models within a single set of p...

k-Neighbor Based Curriculum Sampling for Sequence Prediction

Multi-step ahead prediction in language models is challenging due to the...