English
Related papers

Related papers: Why Transformers Need Adam: A Hessian Perspective

200 papers

Stochastic gradient descent (SGD) optimization methods are nowadays the method of choice for the training of deep neural networks (DNNs) in artificial intelligence systems. In practically relevant training problems, usually not the plain…

Optimization and Control · Mathematics 2024-08-01 Steffen Dereich , Arnulf Jentzen

Stochastic gradient methods (SGMs) are the predominant approaches to train deep learning models. The adaptive versions (e.g., Adam and AMSGrad) have been extensively used in practice, partly because they achieve faster convergence than the…

Optimization and Control · Mathematics 2022-04-14 Yangyang Xu , Yibo Xu , Yonggui Yan , Colin Sutcher-Shepard , Leopold Grinberg , Jie Chen

Models are expected to engage in invariance learning, which involves distinguishing the core relations that remain consistent across varying environments to ensure the predictions are safe, robust and fair. While existing works consider…

Machine Learning · Computer Science 2025-03-11 Yang Xu , Yihong Gu , Cong Fang

Behavioral homogeneity is often critical for the functioning of network systems of interacting entities. In power grids, whose stable operation requires generator frequencies to be synchronized--and thus homogeneous--across the network,…

Adaptation and Self-Organizing Systems · Physics 2024-09-24 Ferenc Molnar , Takashi Nishikawa , Adilson E. Motter

1-bit gradient compression and local steps are two representative techniques that enable drastic communication reduction in distributed SGD. Their benefits, however, remain an open question on Adam-based large model pre-training (e.g. BERT…

Machine Learning · Computer Science 2022-05-24 Yucheng Lu , Conglong Li , Minjia Zhang , Christopher De Sa , Yuxiong He

Transformer becomes the state-of-the-art translation model, while it is not well studied how each intermediate component contributes to the model performance, which poses significant challenges for designing optimal architectures. In this…

Computation and Language · Computer Science 2020-11-10 Wenxuan Wang , Zhaopeng Tu

Which parts of a dataset will a given model find difficult? Recent work has shown that SGD-trained models have a bias towards simplicity, leading them to prioritize learning a majority class, or to rely upon harmful spurious correlations.…

Machine Learning · Computer Science 2023-06-09 Samuel J. Bell , Levent Sagun

Since its inception in "Attention Is All You Need", transformer architecture has led to revolutionary advancements in NLP. The attention layer within the transformer admits a sequence of input tokens $X$ and makes them interact through…

Machine Learning · Computer Science 2024-02-23 Davoud Ataee Tarzanagh , Yingcong Li , Christos Thrampoulidis , Samet Oymak

Federated learning is an emerging research paradigm enabling collaborative training of machine learning models among different organizations while keeping data private at each institution. Despite recent progress, there remain fundamental…

Machine Learning · Computer Science 2022-04-15 Liangqiong Qu , Yuyin Zhou , Paul Pu Liang , Yingda Xia , Feifei Wang , Ehsan Adeli , Li Fei-Fei , Daniel Rubin

Stochastic Gradient Decent (SGD) is one of the core techniques behind the success of deep neural networks. The gradient provides information on the direction in which a function has the steepest rate of change. The main problem with basic…

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, showing that Adam's…

Machine Learning · Computer Science 2020-05-06 Wenjie Li , Zhaoyang Zhang , Xinjiang Wang , Ping Luo

Machine learning models trained with \emph{stochastic} gradient descent (SGD) can generalize better than those trained with deterministic gradient descent (GD). In this work, we study SGD's impact on generalization through the lens of the…

Machine Learning · Computer Science 2025-12-09 Hongjian Lan , Yucong Liu , Florian Schäfer

We consider the distributed learning problem with data dispersed across multiple workers under the orchestration of a central server. Asynchronous Stochastic Gradient Descent (SGD) has been widely explored in such a setting to reduce the…

Machine Learning · Computer Science 2024-05-28 Xiaolu Wang , Yuchang Sun , Hoi-To Wai , Jun Zhang

Adam is the default algorithm for training neural networks, including large language models (LLMs). However, \citet{reddi2019convergence} provided an example that Adam diverges, raising concerns for its deployment in AI model training. We…

Machine Learning · Computer Science 2026-03-03 Yushun Zhang , Bingran Li , Congliang Chen , Zhi-Quan Luo , Ruoyu Sun

Adaptive gradient methods (AGMs) have become popular in optimizing the nonconvex problems in deep learning area. We revisit AGMs and identify that the adaptive learning rate (A-LR) used by AGMs varies significantly across the dimensions of…

Machine Learning · Computer Science 2019-09-12 Qianqian Tong , Guannan Liang , Jinbo Bi

SGD with momentum is one of the key components for improving the performance of neural networks. For decentralized learning, a straightforward approach using momentum is Distributed SGD (DSGD) with momentum (DSGDm). However, DSGDm performs…

Machine Learning · Computer Science 2023-09-26 Yuki Takezawa , Han Bao , Kenta Niwa , Ryoma Sato , Makoto Yamada

The training of Generative Adversarial Networks is a difficult task mainly due to the nature of the networks. One such issue is when the generator and discriminator start oscillating, rather than converging to a fixed point. Another case…

Machine Learning · Statistics 2018-02-08 Alexey Chaplygin , Joshua Chacksfield

We provide a simple proof of convergence covering both the Adam and Adagrad adaptive optimization algorithms when applied to smooth (possibly non-convex) objective functions with bounded gradients. We show that in expectation, the squared…

Machine Learning · Statistics 2022-10-18 Alexandre Défossez , Léon Bottou , Francis Bach , Nicolas Usunier

Deep learning methods - usually consisting of a class of deep neural networks (DNNs) trained by a stochastic gradient descent (SGD) optimization method - are nowadays omnipresent in data-driven learning problems as well as in scientific…

Optimization and Control · Mathematics 2025-01-13 Steffen Dereich , Arnulf Jentzen , Adrian Riekert

Synchronous federated learning scales poorly due to the straggler effect. Asynchronous algorithms increase the update throughput by processing updates upon arrival, but they introduce two fundamental challenges: gradient staleness, which…

Machine Learning · Computer Science 2026-03-30 Abdelkrim Alahyane , Céline Comte , Matthieu Jonckheere