Implicit Unlikelihood Training: Improving Neural Text Generation with Reinforcement Learning
This addresses text generation issues for NLP applications, but it is incremental as it builds on existing unlikelihood training methods.
The paper tackles the problem of dull and repetitive text generation from language models by proposing a reinforcement learning fine-tuning approach that minimizes repetition, showing that it reduces repetition without harming model quality.
Likelihood training and maximization-based decoding result in dull and repetitive generated texts even when using powerful language models (Holtzman et al., 2019). Adding a loss function for regularization was shown to improve text generation output by helping avoid unwanted properties, such as contradiction or repetition (Li at al., 2020). In this work, we propose fine-tuning a language model by using policy gradient reinforcement learning, directly optimizing for better generation. We apply this approach to minimizing repetition in generated text, and show that, when combined with unlikelihood training (Welleck et al., 2020), our method further reduces repetition without impacting the language model quality. We also evaluate other methods for improving generation at training and decoding time, and compare them using various metrics aimed at control for better text generation output.