|
ИСТИНА |
Войти в систему Регистрация |
ФНКЦ РР |
||
Survey on Efficient Training of Large Neural Networksстатья
- Авторы: Gusak Julia, Cherniuk Daria, Shilova Alena, Katrutsa Alexandr, Bershatsky Daniel, Zhao Xunyi, Eyraud-Dubois Lionel, Shliazhko Oleh, Dimitrov Denis, Oseledets Ivan, Beaumont Olivier
- Сборник: Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence
- Год издания: 2022
- Место издания: International Joint Conferences on Artificial Intelligence Organization
- Первая страница: 5494
- Последняя страница: 5501
- DOI: 10.24963/ijcai.2022/769
- Аннотация: Modern Deep Neural Networks (DNNs) require significant memory to store weight, activations, and other intermediate tensors during training. Hence, many models don’t fit one GPU device or can be trained using only a small per-GPU batch size. This survey provides a systematic overview of the approaches that enable more efficient DNNs training. We analyze techniques that save memory and make good use of computation and communication resources on architectures with a single or several GPUs. We summarize the main categories of strategies and compare strategies within and across categories. Along with approaches proposed in the literature, we discuss available implementations.
- Добавил в систему: Димитров Денис Валерьевич
Прикрепленные файлы
| № | Имя | Описание | Имя файла | Размер | Добавлен |
|---|---|---|---|---|---|
| 1. | Полный текст | Полный текст статьи | Survey_on_Efficient_Training_of_Large_Neural_Networks.pdf | 158,4 КБ | 27 августа 2022 [DimitrovDV] |