Memory-Efficient Optimization with Factorized Hamiltonian Descent
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866916442417397760 |
|---|---|
| author | Nguyen, Son Chen, Lizhang Liu, Bo Liu, Qiang |
| author_facet | Nguyen, Son Chen, Lizhang Liu, Bo Liu, Qiang |
| contents | Modern deep learning heavily depends on adaptive optimizers such as Adam and its variants, which are renowned for their capacity to handle model scaling and streamline hyperparameter tuning. However, these algorithms typically experience high memory overhead caused by the accumulation of optimization states, leading to a critical challenge in training large-scale network models. In this study, we introduce a novel adaptive optimizer, H-Fac, which incorporates a memory-efficient factorization approach to address this challenge. By employing a rank-1 parameterization for both momentum and scaling parameter estimators, H-Fac reduces memory costs to a sublinear level while maintaining competitive performance across a wide range of architectures. We develop our algorithms based on principles derived from Hamiltonian dynamics, providing robust theoretical underpinnings in optimization dynamics and convergence guarantees. These optimization algorithms are designed to be both straightforward and adaptable, facilitating easy implementation in diverse settings. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_09958 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Memory-Efficient Optimization with Factorized Hamiltonian Descent Nguyen, Son Chen, Lizhang Liu, Bo Liu, Qiang Machine Learning Modern deep learning heavily depends on adaptive optimizers such as Adam and its variants, which are renowned for their capacity to handle model scaling and streamline hyperparameter tuning. However, these algorithms typically experience high memory overhead caused by the accumulation of optimization states, leading to a critical challenge in training large-scale network models. In this study, we introduce a novel adaptive optimizer, H-Fac, which incorporates a memory-efficient factorization approach to address this challenge. By employing a rank-1 parameterization for both momentum and scaling parameter estimators, H-Fac reduces memory costs to a sublinear level while maintaining competitive performance across a wide range of architectures. We develop our algorithms based on principles derived from Hamiltonian dynamics, providing robust theoretical underpinnings in optimization dynamics and convergence guarantees. These optimization algorithms are designed to be both straightforward and adaptable, facilitating easy implementation in diverse settings. |
| title | Memory-Efficient Optimization with Factorized Hamiltonian Descent |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2406.09958 |