SageMaker AI 无服务器定制新增 Full Fine-Tuning 支持
Full Fine-Tuning (FFT) on SageMaker AI Serverless Model Customization
AWS 给 SageMaker 无服务器微调加上了全参数微调选项,不用自己管 GPU 也能跑 FFT。博客里把 FFT 和 LoRA 在基准上的差异讲得比较透,还给了超参数参考。经常做微调、纠结 LoRA 够不够用的工程师可以照着流程试一遍。
AWS 宣布 Amazon SageMaker AI 的 Serverless Model Customization 支持 Full Fine-Tuning(FFT),用户可以在无服务器模式下对模型进行全参数微调,无需自行管理 GPU 实例。
官方博客对比了 FFT 与 LoRA 在多个基准上的表现,指出在部分任务上 FFT 的质量优于 LoRA,并从模型权重矩阵的谱特性角度解释了两者差距的来源。文章还给出了实际可用的超参数调优指南,包括学习率等关键设置,并演示了从零开始的操作流程。
对需要在 LoRA 之外追求更高微调质量的团队来说,这条无服务器路径降低了 FFT 的基础设施门槛。开发者可以直接在 SageMaker 控制台或 API 中启用 FFT 作业。
Full Fine-Tuning (FFT) on SageMaker AI Serverless Model Customization
Amazon SageMaker AI Serverless Model Customization now supports Full Fine-Tuning (FFT). This post explores when FFT outperforms LoRA across key benchmarks, examines the spectral properties of model weight matrices behind the quality gap, shares practical hyperparameter tuning guidelines, and shows how to get started.
This report is categorized under AITOP's "How-to" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include SageMaker, AWS, Full Fine-Tuning, LoRA.
Primary source verification: published directly by AWS Machine Learning Blog with first-party authority.