English
Related papers

Related papers: Quantifying Hyperparameter Transfer and the Import…

200 papers

We study empirical scaling laws for transfer learning between distributions in an unsupervised, fine-tuning setting. When we train increasingly large neural networks from-scratch on a fixed-size dataset, they eventually become data-limited…

Machine Learning · Computer Science 2021-02-03 Danny Hernandez , Jared Kaplan , Tom Henighan , Sam McCandlish

Despite its popularity, several recent works question the effectiveness of MAML when test tasks are different from training tasks, thus suggesting various task-conditioned methodology to improve the initialization. Instead of searching for…

Machine Learning · Computer Science 2020-12-09 Sungyong Baik , Myungsub Choi , Janghoon Choi , Heewon Kim , Kyoung Mu Lee

Pretraining large language models is a costly process. To make this process more efficient, several methods have been proposed to optimize model architecture/parametrization and hardware use. On the parametrization side, $\mu P$ (Maximal…

Machine Learning · Computer Science 2025-06-19 Soufiane Hayou , Liyuan Liu

Large language models (LLMs) demonstrate strong performance as text embedding models when finetuned with supervised contrastive training. However, their large size balloons inference time and memory requirements. In this paper, we show that…

Computation and Language · Computer Science 2024-10-21 Thennal D K , Tim Fischer , Chris Biemann

One of the main challenges in optimal scaling of large language models (LLMs) is the prohibitive cost of hyperparameter tuning, particularly learning rate $\eta$ and batch size $B$. While techniques like $\mu$P (Yang et al., 2022) provide…

Machine Learning · Computer Science 2025-01-10 Oleg Filatov , Jan Ebert , Jiangtao Wang , Stefan Kesselheim

The increasing capabilities of Machine Learning (ML) models go hand in hand with an immense amount of data and computational power required for training. Therefore, training is usually outsourced into HPC facilities, where we have started…

Machine Learning · Computer Science 2025-01-28 Sabrina Herbst , Vincenzo De Maio , Ivona Brandic

We provide the first proof of learning rate transfer with width in a linear multi-layer perceptron (MLP) parametrized with $\mu$P, a neural network parameterization designed to ``maximize'' feature learning in the infinite-width limit. We…

Machine Learning · Statistics 2026-02-26 Soufiane Hayou

Multilingual machine translation suffers from negative interference across languages. A common solution is to relax parameter sharing with language-specific modules like adapters. However, adapters of related languages are unable to…

Computation and Language · Computer Science 2022-12-06 Christos Baziotis , Mikel Artetxe , James Cross , Shruti Bhosale

Low-Rank Adaptation (LoRA) is a standard tool for parameter-efficient finetuning of large models. While it induces a small memory footprint, its training dynamics can be surprisingly complex as they depend on several hyperparameters such as…

Machine Learning · Computer Science 2026-02-09 Nan Chen , Soledad Villar , Soufiane Hayou

The cost of hyperparameter tuning in deep learning has been rising with model sizes, prompting practitioners to find new tuning methods using a proxy of smaller networks. One such proposal uses $\mu$P parameterized networks, where the…

Machine Learning · Statistics 2023-12-11 Blake Bordelon , Lorenzo Noci , Mufan Bill Li , Boris Hanin , Cengiz Pehlevan

Transformer-based self-supervised models are trained as feature extractors and have empowered many downstream speech tasks to achieve state-of-the-art performance. However, both the training and inference process of these models may…

Computation and Language · Computer Science 2021-05-04 Jinchuan Tian , Rongzhi Gu , Helin Wang , Yuexian Zou

We re-evaluate the standard practice of sharing weights between input and output embeddings in state-of-the-art pre-trained language models. We show that decoupled embeddings provide increased modeling flexibility, allowing us to…

Computation and Language · Computer Science 2020-10-27 Hyung Won Chung , Thibault Févry , Henry Tsai , Melvin Johnson , Sebastian Ruder

Low-rank adapters enable fine-tuning of large models with only a small number of parameters, thus reducing storage costs and minimizing the risk of catastrophic forgetting. However, they often pose optimization challenges, with poor…

Machine Learning · Computer Science 2024-12-16 Piotr Teterwak , Kate Saenko , Bryan A. Plummer , Ser-Nam Lim

Transfer learning is a useful technique for achieving improved performance and reducing training costs by leveraging the knowledge gained from source tasks and applying it to target tasks. Assessing the effectiveness of transfer learning…

Machine Learning · Computer Science 2023-06-12 Peizhong Ju , Sen Lin , Mark S. Squillante , Yingbin Liang , Ness B. Shroff

The Normalized Transformer, or nGPT (arXiv:2410.01131) achieves impressive training speedups and does not require weight decay or learning rate warmup. However, despite having hyperparameters that explicitly scale with model size, we…

Machine Learning · Computer Science 2026-05-04 Boris Shigida , Boris Hanin , Andrey Gromov

Efficient LLM pre-training requires well-tuned hyperparameters (HPs), including learning rate $\eta$ and weight decay $\lambda$. We study scaling laws for HPs: formulas for how to scale HPs as we scale model size N, dataset size D, and…

Machine Learning · Computer Science 2025-11-25 Shane Bergsma , Nolan Dey , Gurpreet Gosal , Gavia Gray , Daria Soboleva , Joel Hestness

We examine the necessity of interpolation in overparameterized models, that is, when achieving optimal predictive risk in machine learning problems requires (nearly) interpolating the training data. In particular, we consider simple…

Machine Learning · Statistics 2022-06-17 Chen Cheng , John Duchi , Rohith Kuditipudi

Neural network-based emulators for the inference of stellar parameters and elemental abundances represent an increasingly popular methodology in modern spectroscopic surveys. However, these approaches are often constrained by their…

Instrumentation and Methods for Astrophysics · Physics 2025-06-10 Tomasz Różański , Yuan-Sen Ting

Fine-tuning is widely used as the default algorithm for transfer learning from pre-trained models. Parameter inefficiency can however arise when, during transfer learning, all the parameters of a large pre-trained model need to be updated…

Computation and Language · Computer Science 2023-03-07 Yingting Li , Ambuj Mehrish , Shuai Zhao , Rishabh Bhardwaj , Amir Zadeh , Navonil Majumder , Rada Mihalcea , Soujanya Poria

Training a high-quality deep neural network requires choosing suitable hyperparameters, which is a non-trivial and expensive process. Current works try to automatically optimize or design principles of hyperparameters, such that they can…

Machine Learning · Computer Science 2024-02-28 Wuyang Chen , Junru Wu , Zhangyang Wang , Boris Hanin