BLEURT: Learning Robust Metrics for Text Generation

Code

• github.com/google-research/bleurt
  OfficialIn papertf★ 789
• github.com/thu-coai/OpenMEVA
  tf★ 51
• github.com/ShiYaya/Awesome_Evaluation_Metrics_for_Text_Generation
  none★ 6
• github.com/sharanya-dasgupta001/hallushift
  pytorch★ 4

Abstract

Text generation has made significant advances in the last few years. Yet, evaluation metrics have lagged behind, as the most popular choices (e.g., BLEU and ROUGE) may correlate poorly with human judgments. We propose BLEURT, a learned evaluation metric based on BERT that can model human judgments with a few thousand possibly biased training examples. A key aspect of our approach is a novel pre-training scheme that uses millions of synthetic examples to help the model generalize. BLEURT provides state-of-the-art results on the last three years of the WMT Metrics shared task and the WebNLG Competition dataset. In contrast to a vanilla BERT-based approach, it yields superior results even when the training data is scarce and out-of-distribution.

Tasks

Reproductions