Evaluating Parameter Efficient Learning for Generation
This work addresses the efficiency and generalization challenges in adapting large pre-trained models for NLP practitioners, though it is incremental as it builds on existing PERM comparisons.
The paper evaluates parameter efficient learning methods (PERMs) compared to finetuning across sample size, generalization, and faithfulness, finding that PERMs outperform finetuning with fewer samples and achieve up to 6% better faithfulness, and applying Adapter to MT-NLG 530b sets new SOTA on Xsum with ROUGE scores of 49.17, 27.20, and 40.98.
Parameter efficient learning methods (PERMs) have recently gained significant attention as they provide an efficient way for pre-trained language models (PLMs) to adapt to a downstream task. However, these conclusions are mostly drawn from in-domain evaluations over the full training set. In this paper, we present comparisons between PERMs and finetuning from three new perspectives: (1) the effect of sample and model size to in-domain evaluations, (2) generalization to unseen domains and new datasets, and (3) the faithfulness of generations. Our results show that for in-domain settings (a) there is a cross point of sample size for which PERMs will perform better than finetuning when training with fewer samples, and (b) larger PLMs have larger cross points. For cross-domain and cross-dataset cases, we show that (a) Adapter (Houlsby et al., 2019) performs the best amongst all the PERMs studied here, and (b) it outperforms finetuning if the task dataset is below a certain size. We also compare the faithfulness of generations and show that PERMs can achieve better faithfulness score than finetuning, especially for small training set, by as much as 6%. Finally, we apply Adapter to MT-NLG 530b (Smith et al., 2022) and achieve new state-of-the-art results on Xsum (Narayan et al., 2018) for all ROUGE scores (ROUGE-1 49.17, ROUGE-2 27.20, ROUGE-L 40.98).