Switch EMA: A Free Lunch for Better Flatness and Sharpness

• URL: http://arxiv.org/abs/2402.09240v2
• Date: Sun, 06 Oct 2024 17:44:46 GMT
• Title: Switch EMA: A Free Lunch for Better Flatness and Sharpness
• Authors: Siyuan Li, Zicheng Liu, Juanxi Tian, Ge Wang, Zedong Wang, Weiyang Jin, Di Wu, Cheng Tan, Tao Lin, Yang Liu, Baigui Sun, Stan Z. Li,
• Abstract summary: This work unveils the full potential of EMA with a single line of modification, i.e., switching parameters to the original model after each epoch, dubbed as Switch (SEMA) From both theoretical and empirical aspects, we demonstrate that SEMA can help DNNs to reach generalization optima that better trade-off between flatness and sharpness.
• Score: 58.55452862747021
• License:
• Abstract: Exponential Moving Average (EMA) is a widely used weight averaging (WA) regularization to learn flat optima for better generalizations without extra cost in deep neural network (DNN) optimization. Despite achieving better flatness, existing WA methods might fall into worse final performances or require extra test-time computations. This work unveils the full potential of EMA with a single line of modification, i.e., switching the EMA parameters to the original model after each epoch, dubbed as Switch EMA (SEMA). From both theoretical and empirical aspects, we demonstrate that SEMA can help DNNs to reach generalization optima that better trade-off between flatness and sharpness. To verify the effectiveness of SEMA, we conduct comparison experiments with discriminative, generative, and regression tasks on vision and language datasets, including image classification, self-supervised learning, object detection and segmentation, image generation, video prediction, attribute regression, and language modeling. Comprehensive results with popular optimizers and networks show that SEMA is a free lunch for DNN training by improving performances and boosting convergence speeds.

Related papers

• Unlearning as multi-task optimization: A normalized gradient difference approach with an adaptive learning rate [105.86576388991713]
  We introduce a normalized gradient difference (NGDiff) algorithm, enabling us to have better control over the trade-off between the objectives. We provide a theoretical analysis and empirically demonstrate the superior performance of NGDiff among state-of-the-art unlearning methods on the TOFU and MUSE datasets.
  arXiv  Detail & Related papers  (2024-10-29T14:41:44Z)
• CRoFT: Robust Fine-Tuning with Concurrent Optimization for OOD Generalization and Open-Set OOD Detection [42.33618249731874]
  We show that minimizing the magnitude of energy scores on training data leads to domain-consistent Hessians of classification loss. We have developed a unified fine-tuning framework that allows for concurrent optimization of both tasks.
  arXiv  Detail & Related papers  (2024-05-26T03:28:59Z)
• Unleashing Network Potentials for Semantic Scene Completion [50.95486458217653]
  This paper proposes a novel SSC framework - Adrial Modality Modulation Network (AMMNet) AMMNet introduces two core modules: a cross-modal modulation enabling the interdependence of gradient flows between modalities, and a customized adversarial training scheme leveraging dynamic gradient competition. Extensive experimental results demonstrate that AMMNet outperforms state-of-the-art SSC methods by a large margin.
  arXiv  Detail & Related papers  (2024-03-12T11:48:49Z)
• Big Learning Expectation Maximization [13.709094150105566]
  We present the Big Learning EM (BigLearn-EM), an EM upgrade that simultaneously performs joint, marginal, and orthogonally transformed marginal matchings. We empirically show that the BigLearn-EM is capable of delivering the optimal with high probability.
  arXiv  Detail & Related papers  (2023-12-19T08:07:41Z)
• How to Scale Your EMA [20.94711634514331]
  We provide a scaling rule for optimization in the presence of a model EMA. We show the rule's validity where the model EMA contributes to the optimization of the target model. For Self-Supervised Learning, we enable training of BYOL up to batch size 24,576 without sacrificing performance.
  arXiv  Detail & Related papers  (2023-07-25T20:33:48Z)
• Bidirectional Looking with A Novel Double Exponential Moving Average to Adaptive and Non-adaptive Momentum Optimizers [109.52244418498974]
  We propose a novel textscAdmeta (textbfADouble exponential textbfMov averagtextbfE textbfAdaptive and non-adaptive momentum) framework. We provide two implementations, textscAdmetaR and textscAdmetaS, the former based on RAdam and the latter based on SGDM.
  arXiv  Detail & Related papers  (2023-07-02T18:16:06Z)
• Magic ELF: Image Deraining Meets Association Learning and Transformer [63.761812092934576]
  This paper aims to unify CNN and Transformer to take advantage of their learning merits for image deraining. A novel multi-input attention module (MAM) is proposed to associate rain removal and background recovery. Our proposed method (dubbed as ELF) outperforms the state-of-the-art approach (MPRNet) by 0.25 dB on average.
  arXiv  Detail & Related papers  (2022-07-21T12:50:54Z)
• Learning-To-Ensemble by Contextual Rank Aggregation in E-Commerce [8.067201256886733]
  We propose a new Learning-To-Ensemble framework RAEGO, which replaces the ensemble model with a contextual Rank Aggregator. RA-EGO has been deployed in our online system and has improved the revenue significantly.
  arXiv  Detail & Related papers  (2021-07-19T03:24:06Z)
• When Vision Transformers Outperform ResNets without Pretraining or Strong Data Augmentations [111.44860506703307]
  Vision Transformers (ViTs) and existing VisionNets signal efforts on replacing hand-wired features or inductive throughputs with general-purpose neural architectures. This paper investigates ViTs and Res-Mixers from the lens of loss geometry, intending to improve the models' data efficiency at training and inference. We show that the improved robustness attributes to sparser active neurons in the first few layers. The resultant ViTs outperform Nets of similar size and smoothness when trained from scratch on ImageNet without large-scale pretraining or strong data augmentations.
  arXiv  Detail & Related papers  (2021-06-03T02:08:03Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.