PulseAugur
中
实时 08:51:02

新的WIS-MeZO方法提高了LLM微调效率

作者 PulseAugur 编辑部 · [2 个来源] ·

研究人员开发了一种名为Wanda-Importance Sparsity-MeZO (WIS-MeZO) 的新方法，以提高大型语言模型零阶（ZO）优化的效率。该技术利用Wanda权重-激活显著性来创建任务特定的稀疏扰动掩码，从而减小搜索空间的维度。在LLaMA和Qwen模型上针对16个基准测试进行了测试，WIS-MeZO与现有的MeZO以及随机或基于幅度的稀疏扰动方法相比，显示出更优的总体得分和更快的收敛速度。 AI

影响 该方法可能导致更高效的大型语言模型微调，降低计算成本和时间。

排序理由 该集群描述了一篇详细介绍优化大型语言模型新方法的最新研究论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 Hugging Face Daily Papers 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →

新的WIS-MeZO方法提高了LLM微调效率

本文如何被排名

Signal score
1 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
该集群描述了一篇详细介绍优化大型语言模型新方法的最新研究论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
2 independent sources
Multiple independent publishers reporting the same story raises confidence that it's real and newsworthy.
Topics
paper, model release
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
2 days old
Coverage has settled into its steady-state source set.
Coverage growth since scoring
+1 source(s) since last score
New sources have picked up this story since our last re-score. Score will update on the next scoring pass.

完整方法见我们的编辑标准。

报道来源 [2]

  1. arXiv cs.AI TIER_1 English(EN) · Ziqing Wen, Xiaohui Song, Jiahuan Wang, Ping Luo, Junlin Zeng, Dongsheng Li, Tao Sun ·

    Wanda引导的稀疏扰动用于高效的零阶优化

    arXiv:2611.16317v1 Announce Type: cross Abstract: Zeroth-order (ZO) optimization enables memory-efficient fine-tuning of large language models through forward-only evaluations, but high-dimensional random perturbations can limit optimization efficiency. Sparse perturbations reduc…

  2. Hugging Face Daily Papers TIER_1 English(EN) ·

    Wanda 指导的稀疏扰动用于高效的零阶优化

    Zeroth-order (ZO) optimization enables memory-efficient fine-tuning of large language models through forward-only evaluations, but high-dimensional random perturbations can limit optimization efficiency. Sparse perturbations reduce the search dimension, making the selection of in…

相关实体

相关话题