两篇新研究论文探讨了提高Transformer模型效率的方法，特别是在边缘设备上部署方面。第一篇论文介绍了OrpQuant，一个无乘法器、二的幂量化的框架，将LLaMA-2-7B等模型的校准时间缩短至约15分钟。第二篇论文研究了残差自由Transformer，证明它们通过保持近乎高斯激活，比传统残差模型对低比特量化表现出更强的鲁棒性。 AI
影响 这些架构和量化创新可以显著降低在资源受限设备上部署大型Transformer模型的计算和内存需求。
排序理由 两篇发表在arXiv上的学术论文，详细介绍了Transformer模型量化和架构设计的新颖方法。
AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →
arXiv:2605.26092v1 Announce Type: cross Abstract: The deployment of Large Language Models (LLMs) and Vision Transformers (ViTs) on edge devices is significantly constrained by memory limitations and the critical timing bottlenecks introduced by dense Multiply-Accumulate (MAC) arr…
arXiv:2605.25880v1 Announce Type: new Abstract: Large-scale transformer training and deployment are increasingly constrained by the transfer of activations, gradients, and optimizer states across accelerators. Low-bit quantization offers a natural remedy, but transformer activati…
The deployment of Large Language Models (LLMs) and Vision Transformers (ViTs) on edge devices is significantly constrained by memory limitations and the critical timing bottlenecks introduced by dense Multiply-Accumulate (MAC) arrays. In the ultra-low bit regime, logarithmic Powe…
Large-scale transformer training and deployment are increasingly constrained by the transfer of activations, gradients, and optimizer states across accelerators. Low-bit quantization offers a natural remedy, but transformer activations are often heavy-tailed and outlier-dominated…