Zobrazeno 1 - 10
of 2 566
pro vyhledávání: '"Zhang Wensheng"'
Different from the traditional semi-supervised learning paradigm that is constrained by the close-world assumption, Generalized Category Discovery (GCD) presumes that the unlabeled dataset contains new categories not appearing in the labeled set, and
Externí odkaz:
http://arxiv.org/abs/2410.21705
This paper explores concurrent FL processes within a three-tier system, with edge servers between edge devices and FL servers. A challenge in this setup is the limited bandwidth from edge devices to edge servers. Thus, allocating the bandwidth effici
Externí odkaz:
http://arxiv.org/abs/2409.04921
Parameter-efficient tunings (PETs) have demonstrated impressive performance and promising perspectives in training large models, while they are still confronted with a common problem: the trade-off between learning new content and protecting old know
Externí odkaz:
http://arxiv.org/abs/2405.13383
Large language models (LLMs) show excellent performance in difficult tasks, but they often require massive memories and computational resources. How to reduce the parameter scale of LLMs has become research hotspots. In this study, we make an importa
Externí odkaz:
http://arxiv.org/abs/2404.09695
Autoregressive decoding strategy is a commonly used method for text generation tasks with pre-trained language models, while early-exiting is an effective approach to speedup the inference stage. In this work, we propose a novel decoding strategy nam
Externí odkaz:
http://arxiv.org/abs/2403.14919
Autor:
Huang, Huimin, Zhang, Wensheng
This paper investigates the convergence rate for Tikhonov regularization of the problem of identifying the coefficient $a \in L^{\infty}(\Omega)$ in the Robin-boundary equation $-\mathrm{div}(a\nabla u)-bu=f,~ x \in \Omega \subset \mathbb R^M,~ M \ge
Externí odkaz:
http://arxiv.org/abs/2403.10229
Publikováno v:
E3S Web of Conferences, Vol 356, p 01045 (2022)
A kind of phase change heat storage electric heating modules filled with the composite PCM was designed and fabricated in this paper. The thermal performance of the module was studied through the experiments, and the thermal performance influencing f
Externí odkaz:
https://doaj.org/article/cdaba8ce30ae4cbaaece7ecd5a60a890
Autor:
Liu, Pinglan, Zhang, Wensheng
With the popularity of cloud computing and machine learning, it has been a trend to outsource machine learning processes (including model training and model-based inference) to cloud. By the outsourcing, other than utilizing the extensive and scalabl
Externí odkaz:
http://arxiv.org/abs/2308.00963
The increasing size of language models raises great research interests in parameter-efficient fine-tuning such as LoRA that freezes the pre-trained model, and injects small-scale trainable parameters for multiple downstream tasks (e.g., summarization
Externí odkaz:
http://arxiv.org/abs/2305.08285
Self-supervised skeleton-based action recognition enjoys a rapid growth along with the development of contrastive learning. The existing methods rely on imposing invariance to augmentations of 3D skeleton within a single data stream, which merely lev
Externí odkaz:
http://arxiv.org/abs/2305.02324