Rethinking Weight Decay for Robust Fine-Tuning of Foundation Models

• URL: http://arxiv.org/abs/2411.01713v1
• Date: Sun, 03 Nov 2024 23:36:53 GMT
• Title: Rethinking Weight Decay for Robust Fine-Tuning of Foundation Models
• Authors: Junjiao Tian, Chengyue Huang, Zsolt Kira,
• Abstract summary: This paper proposes a new weight decay technique, Selective Projection Decay (SPD) SPD selectively imposes a strong penalty on certain layers while allowing others to change freely. When equipped with SPD, Adam consistently provides better in-distribution robustness and out-of-distribution performance on benchmarks.
• Score: 27.847140934456288
• License:
• Abstract: Modern optimizers such as AdamW, equipped with momentum and adaptive learning rate, are designed to escape local minima and explore the vast parameter space. This exploration is beneficial for finding good loss basins when training from scratch. It is not necessarily ideal when resuming from a powerful foundation model because it can lead to large deviations from the pre-trained initialization and, consequently, worse robustness and generalization. At the same time, strong regularization on all parameters can lead to under-fitting. We hypothesize that selectively regularizing the parameter space is the key to fitting and retraining the pre-trained knowledge. This paper proposes a new weight decay technique, Selective Projection Decay (SPD), that selectively imposes a strong penalty on certain layers while allowing others to change freely. Intuitively, SPD expands and contracts the parameter search space for layers with consistent and inconsistent loss reduction, respectively. Experimentally, when equipped with SPD, Adam consistently provides better in-distribution generalization and out-of-distribution robustness performance on multiple popular vision and language benchmarks. Code available at~\url{https://github.com/GT-RIPL/Selective-Projection-Decay.git}

Related papers

• Sparse is Enough in Fine-tuning Pre-trained Large Language Models [98.46493578509039]
  We propose a gradient-based sparse fine-tuning algorithm, named Sparse Increment Fine-Tuning (SIFT) We validate its effectiveness on a range of tasks including the GLUE Benchmark and Instruction-tuning.
  arXiv  Detail & Related papers  (2023-12-19T06:06:30Z)
• FedNAR: Federated Optimization with Normalized Annealing Regularization [54.42032094044368]
  We explore the choices of weight decay and identify that weight decay value appreciably influences the convergence of existing FL algorithms. We develop Federated optimization with Normalized Annealing Regularization (FedNAR), a plug-in that can be seamlessly integrated into any existing FL algorithms.
  arXiv  Detail & Related papers  (2023-10-04T21:11:40Z)
• PathProx: A Proximal Gradient Algorithm for Weight Decay Regularized Deep Neural Networks [25.114642281756495]
  Weight decay is one of the most widely used forms of regularization in deep learning. This paper argues that gradient descent may be an inefficient algorithm for this objective. For neural networks with ReLU activations, solutions to the weight decay objective are equivalent to those of a different objective.
  arXiv  Detail & Related papers  (2022-10-06T17:22:40Z)
• Adaptive Self-supervision Algorithms for Physics-informed Neural Networks [59.822151945132525]
  Physics-informed neural networks (PINNs) incorporate physical knowledge from the problem domain as a soft constraint on the loss function. We study the impact of the location of the collocation points on the trainability of these models. We propose a novel adaptive collocation scheme which progressively allocates more collocation points to areas where the model is making higher errors.
  arXiv  Detail & Related papers  (2022-07-08T18:17:06Z)
• Robust Training of Neural Networks using Scale Invariant Architectures [70.67803417918854]
  In contrast to SGD, adaptive gradient methods like Adam allow robust training of modern deep networks. We show that this general approach is robust to rescaling of parameter and loss. We design a scale invariant version of BERT, called SIBERT, which when trained simply by vanilla SGD achieves performance comparable to BERT trained by adaptive methods like Adam.
  arXiv  Detail & Related papers  (2022-02-02T11:58:56Z)
• Test-time Batch Statistics Calibration for Covariate Shift [66.7044675981449]
  We propose to adapt the deep models to the novel environment during inference. We present a general formulation $alpha$-BN to calibrate the batch statistics. We also present a novel loss function to form a unified test time adaptation framework Core.
  arXiv  Detail & Related papers  (2021-10-06T08:45:03Z)
• The Implicit Biases of Stochastic Gradient Descent on Deep Neural Networks with Batch Normalization [44.30960913470372]
  Deep neural networks with batch normalization (BN-DNNs) are invariant to weight rescaling due to their normalization operations. We investigate the implicit biases of gradient descent (SGD) on BN-DNNs to provide a theoretical explanation for the efficacy of weight decay.
  arXiv  Detail & Related papers  (2021-02-06T03:40:20Z)
• AdamP: Slowing Down the Slowdown for Momentum Optimizers on Scale-invariant Weights [53.8489656709356]
  Normalization techniques are a boon for modern deep learning. It is often overlooked, however, that the additional introduction of momentum results in a rapid reduction in effective step sizes for scale-invariant weights. In this paper, we verify that the widely-adopted combination of the two ingredients lead to the premature decay of effective step sizes and sub-optimal model performances.
  arXiv  Detail & Related papers  (2020-06-15T08:35:15Z)
• A Batch Normalized Inference Network Keeps the KL Vanishing Away [35.40781000297285]
  Variational Autoencoder (VAE) is widely used to approximate a model's posterior on latent variables. VAE often converges to a degenerated local optimum known as "posterior collapse"
  arXiv  Detail & Related papers  (2020-04-27T05:20:01Z)

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.