English(EN) Your LR Schedule May Not Be Principled: Effective Learning Rate Scheduling in Weight-Normalized Training

新的AIR学习率调度优化了权重归一化的AI训练

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-29 00:00

一篇新论文介绍了一种名为AIR（仿射平方根倒数衰减）的学习率调度方法，旨在优化权重归一化训练。该方法侧重于控制有效学习率（ELR），而非仅控制标称学习率，后者已被证明是权重归一化训练中获得提升的主要驱动因素。在1280亿token的混合专家模型预训练实验中，AIR的表现优于传统的余弦调度，从而降低了训练损失并提高了下游性能。 AI

影响这项研究可能导致更高效、更鲁棒的大型AI模型训练，从而降低计算成本并提高性能。

排序理由该集群包含一篇研究论文，详细介绍了一种优化AI模型训练的新方法。[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 1 个来源。我们如何撰写摘要 →

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; 该集群包含一篇研究论文，详细介绍了一种优化AI模型训练的新方法。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, infra; 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: 1 days old; Coverage has settled into its steady-state source set.

完整方法见我们的编辑标准。

报道来源 [1]

Hugging Face Daily Papers TIER_1 English(EN) · 2026-10-29 00:00

您的LR调度可能并非基于原则：权重归一化训练中的有效学习率调度

Effective learning rate trajectories primarily drive weight-normalized training gains, motivating an affine inverse-square-root schedule that improves large-scale pretraining loss and robustness.

本文如何被排名

报道来源 [1]

您的LR调度可能并非基于原则：权重归一化训练中的有效学习率调度

相关实体

相关话题