paper-with-me

홈 › Papers

ParaGAN: A Scalable Distributed Training Framework for Generative Adversarial Networks

2024-11-06 · Ziji Shi, Jialin Li, Yang You

Recent advances in Generative Artificial Intelligence have fueled numerous applications, particularly those involving Generative Adversarial Networks (GANs), which are essential for synthesizing realistic photos and videos. However, efficiently training GANs remains a critical challenge due to their computationally intensive and numerically unstable nature. Existing methods often require days or even weeks for training, posing significant resource and time constraints. In this work, we introduce ParaGAN, a scalable distributed GAN training framework that leverages asynchronous training and an asymmetric optimization policy to accelerate GAN training. ParaGAN employs a congestion-aware data pipeline and hardware-aware layout transformation to enhance accelerator utilization, resulting in over 30% improvements in throughput. With ParaGAN, we reduce the training time of BigGAN from 15 days to 14 hours while achieving 91% scaling efficiency. Additionally, ParaGAN enables unprecedented high-resolution image generation using BigGAN.

📄 PDF Abstract BibTeX arXiv:2411.03999

Code (0)

등록된 구현이 없습니다.

Tasks

Image Generation

Methods 이 논문이 사용한 방법론

((Reservation@Faqs))How do I cancel a reservation on Expedia? How do I cancel a reservation on Expedia? +1^888^829^0881° oR +1^888^829^0881 – Need to cancel your Expedia reservation quickly and without hassle? This step-by-step guide…
Six Ways To Communicate To Someone At Expedia Via Phone And Email's. To communicate or get human at Expedia, the quickest option is typically to call their customer service at +1-888-829-0881 or +1(805) 330 (4056). You can also use the live chat…
ReLU How Do I Communicate to Expedia? How Do I Communicate to Expedia? – Call ☎️ +1-(888) 829 (0881) or +1-805-330-4056 or +1-805-330-4056 for Live Support & Special Travel…
Non-Local Operation A Non-Local Operation is a component for capturing long-range dependencies with deep neural networks. It is a generalization of the classical non-local mean operation in…
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Adam 설명 없음
Residual Connection 설명 없음
Early Stopping Early Stopping is a regularization technique for deep neural networks that stops training when parameter updates no longer begin to yield improves on a validation set. In…

Similar Papers 제목 키워드 기반

A Parameterized Generative Adversarial Network Using Cyclic Projection for Explainable Medical Image Classification

2023-11-24 · Xiangyu Xiong, Yue Sun, Xiaohong Liu, Chan-Tong Lam 외

Although current data augmentation methods are successful to alleviate the data insufficiency, conventional augmentation are primarily intra-domain while advanced generative adversarial networks (GANs) generate images re…

Data AugmentationGenerative Adversarial Networkimage-classificationImage Classification+1

Deep learning-based auto-segmentation of paraganglioma for growth monitoring

2024-03-19 · E. M. C. Sijben, J. C. Jansen, M. de Ridder, P. A. N. Bosman 외

Volume measurement of a paraganglioma (a rare neuroendocrine tumor that typically forms along major blood vessels and nerve pathways in the head and neck region) is crucial for monitoring and modeling tumor growth in the…

Deep Learning

Function Class Learning with Genetic Programming: Towards Explainable Meta Learning for Tumor Growth Functionals

2024-02-19 · E. M. C. Sijben, J. C. Jansen, P. A. N. Bosman, T. Alderliesten

Paragangliomas are rare, primarily slow-growing tumors for which the underlying growth pattern is unknown. Therefore, determining the best care for a patient is hard. Currently, if no significant tumor growth is observed…

Meta-Learning

PPGL-Swarm: Integrated Multimodal Risk Stratification and Hereditary Syndrome Detection in Pheochromocytoma and Paraganglioma

2026-03-23 · Zelin Liu, Xiangfu Yu, Jie Huang, Ge Wang 외 arxiv

Pheochromocytomas and paragangliomas (PPGLs) are rare neuroendocrine tumors, of which 15-25% develop metastatic disease with 5-year survival rates reported as low as 34%. PPGL may indicate hereditary syndromes requiring …

Reinforcement Learning

Weakly Supervised Detection of Pheochromocytomas and Paragangliomas in CT

2024-02-12 · David C. Oluigboa, Bikash Santra, Tejas Sudharshan Mathai, Pritam Mukherjee 외

Pheochromocytomas and Paragangliomas (PPGLs) are rare adrenal and extra-adrenal tumors which have the potential to metastasize. For the management of patients with PPGLs, CT is the preferred modality of choice for precis…

Management