paper-with-me

홈 › Papers

Nostalgic Adam: Weighting more of the past gradients when designing the adaptive learning rate

2018-05-19 · Haiwen Huang, Chang Wang, Bin Dong

First-order optimization algorithms have been proven prominent in deep learning. In particular, algorithms such as RMSProp and Adam are extremely popular. However, recent works have pointed out the lack of ``long-term memory" in Adam-like algorithms, which could hamper their performance and lead to divergence. In our study, we observe that there are benefits of weighting more of the past gradients when designing the adaptive learning rate. We therefore propose an algorithm called the Nostalgic Adam (NosAdam) with theoretically guaranteed convergence at the best known convergence rate. NosAdam can be regarded as a fix to the non-convergence issue of Adam in alternative to the recent work of [Reddi et al., 2018]. Our preliminary numerical experiments show that NosAdam is a promising alternative algorithm to Adam. The proofs, code and other supplementary materials can be found in an anonymously shared link.

📄 PDF Abstract BibTeX arXiv:1805.07557

Code (2)

andrehuang/NostalgicAdam-NosAdam 공식 구현 pytorch
andrehuang/NostalgicAdam pytorch

Methods 이 논문이 사용한 방법론

RMSProp RMSProp is an unpublished adaptive learning rate optimizer proposed by Geoff Hinton. The motivation…
Adam 설명 없음

Similar Papers 제목 키워드 기반

Adam revisited: a weighted past gradients perspective

2021-01-01 · Hui Zhong, Zaiyi Chen, Chuan Qin, Zai Huang 외

Adaptive learning rate methods have been successfully applied in many fields, especially in training deep neural networks. Recent results have shown that adaptive methods with exponential increasing weights on squared pa…

The AdEMAMix Optimizer: Better, Faster, Older

2024-09-05 · Matteo Pagliardini, Pierre Ablin, David Grangier

Momentum based optimizers are central to a wide range of machine learning applications. These typically rely on an Exponential Moving Average (EMA) of gradients, which decays exponentially the present contribution of old…

image-classificationImage ClassificationLanguage ModelingLanguage Modelling

On the Convergence of Adam and Beyond

2019-04-19 · ICLR 2018 1 · Sashank J. Reddi, Satyen Kale, Sanjiv Kumar

Several recently proposed stochastic optimization methods that have been successfully used in training deep networks such as RMSProp, Adam, Adadelta, Nadam are based on using gradient updates scaled by square roots of ex…

Stochastic Optimization

Divergence Results and Convergence of a Variance Reduced Version of ADAM

2022-10-11 · Ruiqi Wang, Diego Klabjan

Stochastic optimization algorithms using exponential moving averages of the past gradients, such as ADAM, RMSProp and AdaGrad, have been having great successes in many applications, especially in training deep neural net…

Stochastic Optimization

AdaX: Adaptive Gradient Descent with Exponential Long Term Memory

2020-04-21 · Wenjie Li, Zhaoyang Zhang, Xinjiang Wang, Ping Luo

Although adaptive optimization algorithms such as Adam show fast convergence in many machine learning tasks, this paper identifies a problem of Adam by analyzing its performance in a simple non-convex synthetic problem, …