Differentiable Expected BLEU for Text Generation
Neural text generation models such as recurrent networks are typically trained by maximizing data log-likelihood based on cross entropy. Such training objective shows a discrepancy from test criteria like the BLEU metric. Recent work optimizes expected BLEU under the model distribution using policy gradient, while such algorithm can suffer from high variance and become impractical. In this paper, we propose a new Differentiable Expected BLEU (DEBLEU) objective that permits direct optimization of neural generation models with gradient descent. We leverage the decomposability and sparsity of BLEU, and reformulate it with moderate approximations, making the evaluation of the objective and its gradient efficient, comparable to common cross-entropy loss. We further devise a simple training procedure with ground-truth masking and annealing for stable optimization. Experiments on neural machine translation and image captioning show our method significantly improves over both cross-entropy and policy gradient training.
Code (0)
등록된 구현이 없습니다.
Tasks
Image CaptioningMachine TranslationText GenerationTranslationSimilar Papers 제목 키워드 기반
Differentiable lower bound for expected BLEU score
In natural language processing tasks performance of the models is often measured with some non-differentiable metric, such as BLEU score. To use efficient gradient-based methods for optimization, it is a common workaroun…
reinforcement-learningReinforcement LearningReinforcement Learning (RL)Transfer Reward Learning for Policy Gradient-Based Text Generation
Task-specific scores are often used to optimize for and evaluate the performance of conditional text generation systems. However, such scores are non-differentiable and cannot be used in the standard supervised learning …
Conditional Text GenerationImage CaptioningPolicy Gradient MethodsSemantic Similarity+4SparseGAN: Sparse Generative Adversarial Network for Text Generation
It is still a challenging task to learn a neural text generation model under the framework of generative adversarial networks (GANs) since the entire training process is not differentiable. The existing training strategi…
Generative Adversarial NetworkSentenceText GenerationWord EmbeddingsSequence-to-Sequence Learning for Indonesian Automatic Question Generator
Automatic question generation is defined as the task of automating the creation of question given a various of textual data. Research in automatic question generator (AQG) has been conducted for more than 10 years, mainl…
Question AnsweringQuestion GenerationQuestion-GenerationRecurrent Deep Differentiable Logic Gate Networks
While differentiable logic gates have shown promise in feedforward networks, their application to sequential modeling remains unexplored. This paper presents the first implementation of Recurrent Deep Differentiable Logi…