用模型平均缓解持续学习中的遗忘问题，无需存数据或额外参数。
Soup to go: mitigating forgetting during continual learning with model averaging
- 训练中融合当前模型与早期检查点，动态保留旧知识。
- 在图像和语言任务上表现优于现有方法，且不依赖数据缓存。
- 适合资源有限场景，尤其关注长期记忆保持的研究者。
在持续学习中，任务数据按顺序到达，后续任务微调常导致先前任务性能下降，尤其在不同领域间更明显。如何在低计算开销下缓解灾难性遗忘、保留已学知识？受其他模型合并方法和L2回归启发，本文提出序列微调平均（SFA），在训练过程中将当前模型与早期检查点进行融合。相比主流方法需维护历史任务数据缓冲区或每步施加惩罚，SFA无需存储过往数据或多个参数副本即可达到相当甚至更优效果。实验表明，该方法在图像与语言领域均超越任务算术、TIES合并、WiSE-FT等典型合并技术，以及L2和弹性权重巩固等正则化方法。结果揭示了训练期间部分训练模型合并的潜在优势。
原文摘要 · Abstract (English)
In continual learning, where task data arrives in a sequence, fine-tuning on later tasks will often lead to performance degradation on earlier tasks. This is especially pronounced when these tasks come from diverse domains. In this setting, how can we mitigate catastrophic forgetting of earlier tasks and retain what the model has learned with minimal computational expenses? Inspired by other merging methods, and L2-regression, we propose Sequential Fine-tuning with Averaging (SFA), a method that merges currently training models with earlier checkpoints during the course of training. SOTA approaches typically maintain a data buffer of past tasks or impose a penalty at each gradient step. In contrast, our method achieves comparable results without the need to store past data, or multiple copies of parameters for each gradient step. Furthermore, our method outperforms common merging techniques such as Task Arithmetic, TIES Merging, and WiSE-FT, as well as other penalty methods like L2 and Elastic Weight Consolidation. In turn, our method offers insight into the benefits of merging partially-trained models during training across both image and language domains.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。