Reinforcement Learning based Curriculum Optimization for Neural Machine Translation.

ACL 2019

Reinforcement Learning based Curriculum Optimization for Neural Machine Translation.

Jan 19, 2021
|
49 views
|
Details
Abstract: We consider the problem of making efficient use of heterogeneous training data in neural machine translation (NMT). Specifically, given a training dataset with a sentence-level feature such as noise, we seek an optimal curriculum, or order for presenting examples to the system during training. Our curriculum framework allows examples to appear an arbitrary number of times, and thus generalizes data weighting, filtering, and fine-tuning schemes. Rather than relying on prior knowledge to design a curriculum, we use reinforcement learning to learn one automatically, jointly with the NMT system, in the course of a single training run. We show that this approach can beat uniform and filtering baselines on Paracrawl and WMT English-to-French datasets by up to +3.4 BLEU, and match the performance of a hand-designed, state-of-the-art curriculum. Authors: Gaurav Kumar, George Foster, Colin Cherry, Maxim Krikun (Johns Hopkins, Google AI)

Comments
loading...