Optimal Beam Search for Machine Translation
نویسندگان
چکیده
Beam search is a fast and empirically effective method for translation decoding, but it lacks formal guarantees about search error. We develop a new decoding algorithm that combines the speed of beam search with the optimal certificate property of Lagrangian relaxation, and apply it to phraseand syntax-based translation decoding. The new method is efficient, utilizes standard MT algorithms, and returns an exact solution on the majority of translation examples in our test data. The algorithm is 3.5 times faster than an optimized incremental constraint-based decoder for phrase-based translation and 4 times faster for syntax-based translation.
منابع مشابه
When to Finish? Optimal Beam Search for Neural Text Generation (modulo beam size)
In neural text generation such as neural machine translation, summarization, and image captioning, beam search is widely used to improve the output text quality. However, in the neural generation setting, hypotheses can finish in different steps, which makes it difficult to decide when to end beam search to ensure optimality. We propose a provably optimal beam search algorithm that will always ...
متن کاملGeneralized Cyclic Open Shop Scheduling and a Hybrid Algorithm
In this paper, we first introduce a generalized version of open shop scheduling (OSS), called generalized cyclic open shop scheduling (GCOSS) and then develop a hybrid method of metaheuristic to solve this problem. Open shop scheduling is concerned with processing n jobs on m machines, where each job has exactly m operations and operation i of each job has to be processed on machine i . However...
متن کاملSix Challenges for Neural Machine Translation
We explore six challenges for neural machine translation: domain mismatch, amount of training data, rare words, long sentences, word alignment, and beam search. We show both deficiencies and improvements over the quality of phrasebased statistical machine translation.
متن کاملSingle-Queue Decoding for Neural Machine Translation
Neural machine translation models rely on the beam search algorithm for decoding. In practice, we found that the quality of hypotheses in the search space is negatively affected owing to the fixed beam size. To mitigate this problem, we store all hypotheses in a single priority queue and use a universal score function for hypothesis selection. The proposed algorithm is more flexible as the disc...
متن کامل