Calibrated Language Model Fine-Tuning for In- and Out-of-Distribution Data
arXiv:2010.11506
Abstract
Fine-tuned pre-trained language models can suffer from severe miscalibration for both in-distribution and out-of-distribution (OOD) data due to over-parameterization. To mitigate this issue, we propose a regularized fine-tuning method. Our method introduces two types of regularization for better calibration: (1) On-manifold regularization, which generates pseudo on-manifold samples through interpolation within the data manifold. Augmented training with these pseudo samples imposes a smoothness regularization to improve in-distribution calibration. (2) Off-manifold regularization, which encourages the model to output uniform distributions for pseudo off-manifold samples to address the over-confidence issue for OOD data. Our experiments demonstrate that the proposed method outperforms existing calibration methods for text classification in terms of expectation calibration error, misclassification detection, and OOD detection on six datasets. Our code can be found at https://github.com/Lingkai-Kong/Calibrated-BERT-Fine-Tuning.
arXiv commentEMNLP2020 long paper
References in corpus (10)
- Adam: A Method for Stochastic Optimization
- Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
- On Calibration of Modern Neural Networks
- A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks
- Weight Uncertainty in Neural Networks
- Fast is better than free: Revisiting adversarial training
- Manifold Mixup: Better Representations by Interpolating Hidden States
- HDLTex: Hierarchical Deep Learning for Text Classification
- Multiplicative Normalizing Flows for Variational Bayesian Neural Networks
- Learning to Defend by Learning to Attack