FrugalScore: Learning Cheaper, Lighter and Faster Evaluation Metricsfor Automatic Text Generation
This addresses the need for fast and reliable evaluation metrics in NLP, particularly benefiting researchers and practitioners with limited computational resources, though it is incremental as it builds on existing metrics.
The paper tackles the problem of expensive evaluation metrics for automatic text generation by proposing FrugalScore, which learns cheaper, lighter, and faster versions of metrics like BERTScore and MoverScore, retaining 96.8% of performance on average while running 24 times faster and having 35 times fewer parameters.
Fast and reliable evaluation metrics are key to R&D progress. While traditional natural language generation metrics are fast, they are not very reliable. Conversely, new metrics based on large pretrained language models are much more reliable, but require significant computational resources. In this paper, we propose FrugalScore, an approach to learn a fixed, low cost version of any expensive NLG metric, while retaining most of its original performance. Experiments with BERTScore and MoverScore on summarization and translation show that FrugalScore is on par with the original metrics (and sometimes better), while having several orders of magnitude less parameters and running several times faster. On average over all learned metrics, tasks, and variants, FrugalScore retains 96.8% of the performance, runs 24 times faster, and has 35 times less parameters than the original metrics. We make our trained metrics publicly available, to benefit the entire NLP community and in particular researchers and practitioners with limited resources.