Neural machine translation (NMT) models usually suffer from catastrophic forgetting during continual training where the models tend to gradually forget previously learned knowledge and swing to fit the newly added data which may have a different distribution, e.g. a different domain. Although many methods have been proposed to solve this problem, we cannot get to know what causes this phenomenon yet. Under the background of domain adaptation, we investigate the cause of catastrophic forgetting from the perspectives of modules and parameters (neurons). The investigation on the modules of the NMT model shows that some modules have tight relation with the general-domain knowledge while some other modules are more essential in the domain adaptation. And the investigation on the parameters shows that some parameters are important for both the general-domain and in-domain translation and the great change of them during continual training brings about the performance decline in general-domain. We conduct experiments across different language pairs and domains to ensure the validity and reliability of our findings.
Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所）；University of Chinese Academy of Sciences（中国科学院大学）
Neural networks tend to gradually forget the previously learned knowledge when learning multiple tasks sequentially from dynamic data distributions. This problem is called \textit{catastrophic forgetting}, which is a fundamental challenge in the continual learning of neural networks. In this work, we observe that catastrophic forgetting not only occurs in continual learning but also affects the traditional static training. Neural networks, especially neural machine translation models, suffer from catastrophic forgetting even if they learn from a static training set. To be specific, the final m…
Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所）；University of Chinese Academy of Sciences（中国科学院大学）；Tencent（腾讯）
This paper considers continual learning of large-scale pretrained neural machine translation model without accessing the previous training data or introducing model separation. We argue that the widely used regularization-based methods, which perform multi-objective learning with an auxiliary loss, suffer from the misestimate problem and cannot always achieve a good balance between the previous and new tasks. To solve the problem, we propose a two-stage training method based on the local features of the real loss. We first search low forgetting risk regions, where the model can retain the perf…
CIMeC（心智与脑科学中心）；IT University of Copenhagen（哥本哈根信息技术大学）；University of Amsterdam（阿姆斯特丹大学）；University of Trento（特伦托大学）
We study the issue of catastrophic forgetting in the context of neural multimodal approaches to Visual Question Answering (VQA). Motivated by evidence from psycholinguistics, we devise a set of linguistically-informed VQA tasks, which differ by the types of questions involved (Wh-questions and polar questions). We test what impact task difficulty has on continual learning, and whether the order in which a child acquires question types facilitates computational models. Our results show that dramatic forgetting is at play and that task difficulty and order matter. Two well-known current continua…
Lewis & Clark College（路易斯与克拉克学院）；University of Colorado Colorado Springs（科罗拉多大学科罗拉多斯普林斯分校）
In order for artificial neural networks to begin accurately mimicking biological ones, they must be able to adapt to new exigencies without forgetting what they have learned from previous training. Lifelong learning approaches to artificial neural networks attempt to strive towards this goal, yet have not progressed far enough to be realistically deployed for natural language processing tasks. The proverbial roadblock of catastrophic forgetting still gate-keeps researchers from an adequate lifelong learning model. While efforts are being made to quell catastrophic forgetting, there is a lack o…
Tsinghua University（清华大学）；Microsoft Research（微软研究院）；Georgia Institute of Technology（佐治亚理工学院）；Beijing National Research Center for Information Science and Technology（北京信息科学与技术国家研究中心）
In this paper, we investigate the instability in the standard dense retrieval training, which iterates between model training and hard negative selection using the being-trained model. We show the catastrophic forgetting phenomena behind the training instability, where models learn and forget different negative groups during training iterations. We then propose ANCE-Tele, which accumulates momentum negatives from past iterations and approximates future iterations using lookahead negatives, as "teleportations" along the time axis to smooth the learning process. On web search and OpenQA, ANCE-Te…
FOREVER: Forgetting Curve-Inspired Memory Replay for Language Model Continual Learning
The Hong Kong Polytechnic University（香港理工大学）；ShineMo Ltd., China（中国 ShineMo 公司）；Solar System of OVB, Tencent, China（腾讯 OVB 太阳系中国）；Peking University（北京大学）；University College London（伦敦大学学院）；University of Illinois Chicago（伊利诺伊大学芝加哥分校）
An Empirical Study of Catastrophic Forgetting in Large Language Models During Continual Fine-tuning
Pattern Recognition Center, WeChat AI, Tencent Inc（腾讯公司微信AI模式识别中心）；School of Engineering, Westlake University（西湖大学工程学院）；Institute of Advanced Technology, Westlake Institute for Advanced Study（西湖高等研究院先进技术研究所）
A desirable dialog system should be able to continually learn new skills without forgetting old ones, and thereby adapt to new domains or tasks in its life cycle. However, continually training a model often leads to a well-known catastrophic forgetting issue. In this paper, we present Continual Prompt Tuning, a parameter-efficient framework that not only avoids forgetting but also enables knowledge transfer between tasks. To avoid forgetting, we only learn and store a few prompt tokens' embeddings for each task while freezing the backbone pre-trained model. To achieve bi-directional knowledge …
When Forgetting is not Catastrophic: On the Mechanics of Spurious Forgetting
MPI for Intelligent Systems, Tübingen（马克斯·普朗克智能系统研究所（蒂宾根））；Jinesis Lab, University of Toronto & Vector Institute（Jinesis实验室，多伦多大学与向量研究所）；EuroSafeAI；Hector Foundation（赫克托基金会）；Stanford University（斯坦福大学）；ELLIS Institute Tübingen（ELLIS研究所（蒂宾根））
IBM Research Australia（IBM澳大利亚研究院）；University of Melbourne（墨尔本大学）
The creation of large-scale open domain reading comprehension data sets in recent years has enabled the development of end-to-end neural comprehension models with promising results. To use these models for domains with limited training data, one of the most effective approach is to first pretrain them on large out-of-domain source data and then fine-tune them with the limited target data. The caveat of this is that after fine-tuning the comprehension models tend to perform poorly in the source domain, a phenomenon known as catastrophic forgetting. In this paper, we explore methods that overcom…
University of Sussex（萨塞克斯大学）；Naver Labs Europe（Naver欧洲实验室）
Machine learning is shifting towards general-purpose pretrained generative models, trained in a self-supervised manner on large amounts of data, which can then be applied to solve a large number of tasks. However, due to their generic training methodology, these models often fail to meet some of the downstream requirements (e.g., hallucinations in abstractive summarization or style violations in code generation). This raises the important question of how to adapt pre-trained generative models to meet all requirements without destroying their general capabilities ("catastrophic forgetting"). Re…
Multi-task learning is useful in NLP because it is often practically desirable to have a single model that works across a range of tasks. In the medical domain, sequential training on tasks may sometimes be the only way to train models, either because access to the original (potentially sensitive) data is no longer available, or simply owing to the computational costs inherent to joint retraining. A major issue inherent to sequential learning, however, is catastrophic forgetting, i.e., a substantial drop in accuracy on prior tasks when a model is updated for a new task. Elastic Weight Consolid…
