Zobrazeno 1 - 10
of 33
pro vyhledávání: '"Pan, Xingyuan"'
Bilevel optimization has shown its utility across various machine learning settings, yet most algorithms in practice require second-order information, making it challenging to scale them up. Only recently, a paradigm of first-order algorithms emerged
Externí odkaz:
http://arxiv.org/abs/2406.19976
Large Language Models (LLMs) have demonstrated remarkable abilities in general scenarios. Instruction finetuning empowers them to align with humans in various tasks. Nevertheless, the Diversity and Quality of the instruction data remain two main chal
Externí odkaz:
http://arxiv.org/abs/2405.12915
This report describes our VolcTrans system for the WMT22 shared task on large-scale multilingual machine translation. We participated in the unconstrained track which allows the use of external resources. Our system is a transformerbased multilingual
Externí odkaz:
http://arxiv.org/abs/2210.11599
Publikováno v:
In International Communications in Heat and Mass Transfer September 2024 157
Various natural language processing tasks are structured prediction problems where outputs are constructed with multiple interdependent decisions. Past work has shown that domain knowledge, framed as constraints over the output space, can help improv
Externí odkaz:
http://arxiv.org/abs/2006.01209
Publikováno v:
In Behavioural Brain Research 12 March 2023 441
Autor:
Pan, Xingyuan, Srikumar, Vivek
Predicting structured outputs can be computationally onerous due to the combinatorially large output spaces. In this paper, we focus on reducing the prediction time of a trained black-box structured classifier without losing accuracy. To do so, we tr
Externí odkaz:
http://arxiv.org/abs/1806.04245
Autor:
Duan, Meiyu, Wang, Yueying, Qiao, Ya, Wang, Yangyang, Pan, Xingyuan, Hu, Zhuyu, Ran, Yanyue, Fu, Xian, Fan, Yusi, Huang, Lan, Zhou, Fengfeng
Publikováno v:
In Computers in Biology and Medicine September 2022 148
Autor:
Pan, Xingyuan, Srikumar, Vivek
Rectified Linear Units (ReLUs) have been shown to ameliorate the vanishing gradient problem, allow for efficient backpropagation, and empirically promote sparsity in the learned parameters. They have led to state-of-the-art results in a variety of ap
Externí odkaz:
http://arxiv.org/abs/1511.05678
Autor:
Wheeler, Dustin D., Willmering, Matthew M., Sesti, Erika L., Pan, Xingyuan, Saha, Dipta, Stanton, Christopher J., Hayes, Sophia E.
Publikováno v:
In Journal of Magnetic Resonance December 2016 273:19-26