On a continuous time model of gradient descent dynamics and instability in deep learning

• URL: http://arxiv.org/abs/2302.01952v3
• Date: Wed, 13 Sep 2023 19:26:40 GMT
• Title: On a continuous time model of gradient descent dynamics and instability in deep learning
• Authors: Mihaela Rosca and Yan Wu and Chongli Qin and Benoit Dherin
• Abstract summary: We propose the principal flow (PF) as a continuous time flow that approximates gradient descent dynamics. The PF sheds light on the recently observed edge of stability phenomena in deep learning. Using our new understanding of instability we propose a learning rate adaptation method which enables us to control the trade-off between training stability and test set evaluation performance.
• Score: 12.20253214080485
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The recipe behind the success of deep learning has been the combination of neural networks and gradient-based optimization. Understanding the behavior of gradient descent however, and particularly its instability, has lagged behind its empirical success. To add to the theoretical tools available to study gradient descent we propose the principal flow (PF), a continuous time flow that approximates gradient descent dynamics. To our knowledge, the PF is the only continuous flow that captures the divergent and oscillatory behaviors of gradient descent, including escaping local minima and saddle points. Through its dependence on the eigendecomposition of the Hessian the PF sheds light on the recently observed edge of stability phenomena in deep learning. Using our new understanding of instability we propose a learning rate adaptation method which enables us to control the trade-off between training stability and test set evaluation performance.

Related papers

• On discretisation drift and smoothness regularisation in neural network training [0.0]
  We aim to make steps towards an improved understanding of deep learning with a focus on optimisation and model regularisation. We start by investigating gradient descent (GD), a discrete-time algorithm at the basis of most popular deep learning optimisation algorithms. We derive novel continuous-time flows that account for discretisation drift. Unlike the NGF, these new flows can be used to describe learning rate specific behaviours of GD, such as training instabilities observed in supervised learning and two-player games. We then translate insights from continuous time into mitigation strategies for unstable GD dynamics, by constructing novel learning rate schedules and regulariser
  arXiv  Detail & Related papers  (2023-10-21T15:21:36Z)
• Investigating the Edge of Stability Phenomenon in Reinforcement Learning [20.631461205889487]
  We explore the edge of stability phenomenon in reinforcement learning (RL) Despite significant differences to supervised learning, the edge of stability phenomenon can be present in off-policy deep RL. Our results suggest that, while neural network structure can lead to optimisation dynamics that transfer between problem domains, certain aspects of deep RL optimisation can differentiate it from domains such as supervised learning.
  arXiv  Detail & Related papers  (2023-07-09T15:46:27Z)
• Beyond the Edge of Stability via Two-step Gradient Updates [49.03389279816152]
  Gradient Descent (GD) is a powerful workhorse of modern machine learning. GD's ability to find local minimisers is only guaranteed for losses with Lipschitz gradients. This work focuses on simple, yet representative, learning problems via analysis of two-step gradient updates.
  arXiv  Detail & Related papers  (2022-06-08T21:32:50Z)
• A Loss Curvature Perspective on Training Instability in Deep Learning [28.70491071044542]
  We study the evolution of the loss Hessian across many classification tasks in order to understand the effect curvature of the loss has on the training dynamics. Inspired by the conditioning perspective, we show that learning rate warmup can improve training stability just as much as batch normalization.
  arXiv  Detail & Related papers  (2021-10-08T20:25:48Z)
• Acceleration via Fractal Learning Rate Schedules [37.878672787331105]
  We show that the learning rate schedule remains notoriously difficult to understand and expensive to tune. We reinterpret an iterative algorithm from the numerical analysis literature as what we call the Chebyshev learning rate schedule for accelerating vanilla gradient descent. We provide some experiments and discussion to challenge current understandings of the "edge of stability" in deep learning.
  arXiv  Detail & Related papers  (2021-03-01T22:52:13Z)
• Gradient Starvation: A Learning Proclivity in Neural Networks [97.02382916372594]
  Gradient Starvation arises when cross-entropy loss is minimized by capturing only a subset of features relevant for the task. This work provides a theoretical explanation for the emergence of such feature imbalance in neural networks.
  arXiv  Detail & Related papers  (2020-11-18T18:52:08Z)
• Training Generative Adversarial Networks by Solving Ordinary Differential Equations [54.23691425062034]
  We study the continuous-time dynamics induced by GAN training. From this perspective, we hypothesise that instabilities in training GANs arise from the integration error. We experimentally verify that well-known ODE solvers (such as Runge-Kutta) can stabilise training.
  arXiv  Detail & Related papers  (2020-10-28T15:23:49Z)
• Improved Analysis of Clipping Algorithms for Non-convex Optimization [19.507750439784605]
  Recently, citetzhang 2019gradient show that clipped (stochastic) Gradient Descent (GD) converges faster than vanilla GD/SGD. Experiments confirm the superiority of clipping-based methods in deep learning tasks.
  arXiv  Detail & Related papers  (2020-10-05T14:36:59Z)
• Reintroducing Straight-Through Estimators as Principled Methods for Stochastic Binary Networks [85.94999581306827]
  Training neural networks with binary weights and activations is a challenging problem due to the lack of gradients and difficulty of optimization over discrete weights. Many successful experimental results have been achieved with empirical straight-through (ST) approaches. At the same time, ST methods can be truly derived as estimators in the binary network (SBN) model with Bernoulli weights.
  arXiv  Detail & Related papers  (2020-06-11T23:58:18Z)
• Disentangling Adaptive Gradient Methods from Learning Rates [65.0397050979662]
  We take a deeper look at how adaptive gradient methods interact with the learning rate schedule. We introduce a "grafting" experiment which decouples an update's magnitude from its direction. We present some empirical and theoretical retrospectives on the generalization of adaptive gradient methods.
  arXiv  Detail & Related papers  (2020-02-26T21:42:49Z)
• The Break-Even Point on Optimization Trajectories of Deep Neural Networks [64.7563588124004]
  We argue for the existence of the "break-even" point on this trajectory. We show that using a large learning rate in the initial phase of training reduces the variance of the gradient. We also show that using a low learning rate results in bad conditioning of the loss surface even for a neural network with batch normalization layers.
  arXiv  Detail & Related papers  (2020-02-21T22:55:51Z)

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.