Zobrazeno 1 - 7
of 7
pro vyhledávání: '"Shi, Ziji"'
As deep learning models continue to increase in size, the memory requirements for training have surged. While high-level techniques like offloading, recomputation, and compression can alleviate memory pressure, they also introduce overheads. However,
Externí odkaz:
http://arxiv.org/abs/2310.19295
Autor:
Zhang, Shiwei, Diao, Lansong, Wang, Siyu, Cao, Zongyan, Gu, Yiliang, Si, Chang, Shi, Ziji, Zheng, Zhen, Wu, Chuan, Lin, Wei
We present Rhino, a system for accelerating tensor programs with automatic parallelization on AI platform for real production environment. It transforms a tensor program written for a single device into an equivalent distributed program that is capab
Externí odkaz:
http://arxiv.org/abs/2302.08141
Autor:
Shi, Ziji, Jiang, Le, Wang, Ang, Zhang, Jie, Jia, Xianyan, Li, Yong, Wu, Chencan, Li, Jialin, Lin, Wei
Model parallelism has become necessary to train large neural networks. However, finding a suitable model parallel schedule for an arbitrary neural network is a non-trivial task due to the exploding search space. In this work, we present a model paral
Externí odkaz:
http://arxiv.org/abs/2302.00247
More transformer blocks with residual connections have recently achieved impressive results on various tasks. To achieve better performance with fewer trainable parameters, recent methods are proposed to go shallower by parameter sharing or model com
Externí odkaz:
http://arxiv.org/abs/2107.11817
Autor:
Jia, Xianyan, Jiang, Le, Wang, Ang, Xiao, Wencong, Shi, Ziji, Zhang, Jie, Li, Xinyuan, Chen, Langshi, Li, Yong, Zheng, Zhen, Liu, Xiaoyong, Lin, Wei
The scaling up of deep neural networks has been demonstrated to be effective in improving model quality, but also encompasses several training challenges in terms of training efficiency, programmability, and resource adaptability. We present Whale, a
Externí odkaz:
http://arxiv.org/abs/2011.09208
Autor:
Shi, Ziji, Ng, Wee Keong
Path planning is important for the autonomy of Unmanned Aerial Vehicle (UAV), especially for scheduling UAV delivery. However, the operating environment of UAVs is usually uncertain and dynamic. Without proper planning, collisions may happen where mu
Externí odkaz:
http://arxiv.org/abs/1805.03358
Publikováno v:
Proceedings of the AAAI Conference on Artificial Intelligence. 36:8779-8787
More transformer blocks with residual connections have recently achieved impressive results on various tasks. To achieve better performance with fewer trainable parameters, recent methods are proposed to go shallower by parameter sharing or model com