Optimizing ML Training with Metagradient Descent

• URL: http://arxiv.org/abs/2503.13751v1
• Date: Mon, 17 Mar 2025 22:18:24 GMT
• Title: Optimizing ML Training with Metagradient Descent
• Authors: Logan Engstrom, Andrew Ilyas, Benjamin Chen, Axel Feldmann, William Moses, Aleksander Madry,
• Abstract summary: We introduce an algorithm for efficiently calculating metagradients -- gradients through model training -- at scale.<n>We then introduce a "smooth model training" framework that enables effective optimization using metagradients.
• Score: 69.89631748402377
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: A major challenge in training large-scale machine learning models is configuring the training process to maximize model performance, i.e., finding the best training setup from a vast design space. In this work, we unlock a gradient-based approach to this problem. We first introduce an algorithm for efficiently calculating metagradients -- gradients through model training -- at scale. We then introduce a "smooth model training" framework that enables effective optimization using metagradients. With metagradient descent (MGD), we greatly improve on existing dataset selection methods, outperform accuracy-degrading data poisoning attacks by an order of magnitude, and automatically find competitive learning rate schedules.

Related papers

• Multiscale Stochastic Gradient Descent: Efficiently Training Convolutional Neural Networks [6.805997961535213]
  Multiscale Gradient Descent (Multiscale-SGD) is a novel optimization approach that exploits coarse-to-fine training strategies to estimate the gradient at a fraction of the cost. We introduce a new class of learnable, scale-independent Mesh-Free Convolutions (MFCs) that ensure consistent gradient behavior across resolutions. Our results establish a new paradigm for the efficient training of deep networks, enabling practical scalability in high-resolution and multiscale learning tasks.
  arXiv  Detail & Related papers  (2025-01-22T09:13:47Z)
• Learning to Generate Gradients for Test-Time Adaptation via Test-Time Training Layers [18.921532965557475]
  Test-time adaptation aims to fine-tune a trained model online using unlabeled testing data.<n>In this optimization process, unsupervised learning objectives like entropy frequently encounter noisy learning signals.<n>We employ a learning-to-optimize approach to automatically learn an entropy generator called Meta Gradient Generator.
  arXiv  Detail & Related papers  (2024-12-22T07:24:09Z)
• MARS: Unleashing the Power of Variance Reduction for Training Large Models [56.47014540413659]
  We propose a unified training framework for deep neural networks.<n>We introduce three instances of MARS that leverage preconditioned gradient optimization.<n>Results indicate that the implementation of MARS consistently outperforms Adam.
  arXiv  Detail & Related papers  (2024-11-15T18:57:39Z)
• AdaRankGrad: Adaptive Gradient-Rank and Moments for Memory-Efficient LLMs Training and Fine-Tuning [9.51289606759621]
  Training and fine-tuning large language models (LLMs) come with challenges related to memory and computational requirements.<n>Various techniques have been developed to tackle these challenges, such as low-rank adaptation (LoRA)<n>We introduce a new method inspired by a phenomenon we formally prove: as training progresses, the rank of the estimated gradient gradually decreases.
  arXiv  Detail & Related papers  (2024-10-23T13:53:26Z)
• Navigating Scaling Laws: Compute Optimality in Adaptive Model Training [39.96209967632896]
  In recent years, the state-of-the-art in deep learning has been dominated by very large models that have been pre-trained on vast amounts of data. We extend the concept of optimality by allowing for an adaptive' model, i.e. a model that can change its shape during training.
  arXiv  Detail & Related papers  (2023-11-06T16:20:28Z)
• Learning Large-scale Neural Fields via Context Pruned Meta-Learning [60.93679437452872]
  We introduce an efficient optimization-based meta-learning technique for large-scale neural field training. We show how gradient re-scaling at meta-test time allows the learning of extremely high-quality neural fields. Our framework is model-agnostic, intuitive, straightforward to implement, and shows significant reconstruction improvements for a wide range of signals.
  arXiv  Detail & Related papers  (2023-02-01T17:32:16Z)
• Model-Agnostic Multitask Fine-tuning for Few-shot Vision-Language Transfer Learning [59.38343286807997]
  We propose Model-Agnostic Multitask Fine-tuning (MAMF) for vision-language models on unseen tasks. Compared with model-agnostic meta-learning (MAML), MAMF discards the bi-level optimization and uses only first-order gradients. We show that MAMF consistently outperforms the classical fine-tuning method for few-shot transfer learning on five benchmark datasets.
  arXiv  Detail & Related papers  (2022-03-09T17:26:53Z)
• Extrapolation for Large-batch Training in Deep Learning [72.61259487233214]
  We show that a host of variations can be covered in a unified framework that we propose. We prove the convergence of this novel scheme and rigorously evaluate its empirical performance on ResNet, LSTM, and Transformer.
  arXiv  Detail & Related papers  (2020-06-10T08:22:41Z)
• Dynamic Scale Training for Object Detection [111.33112051962514]
  We propose a Dynamic Scale Training paradigm (abbreviated as DST) to mitigate scale variation challenge in object detection. Experimental results demonstrate the efficacy of our proposed DST towards scale variation handling. It does not introduce inference overhead and could serve as a free lunch for general detection configurations.
  arXiv  Detail & Related papers  (2020-04-26T16:48:17Z)

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.