Posted 2026-10-09Updated 2026-10-15aia few seconds read (About 44 words)post论文阅读随记，施工中 QLoRA通过以下几个点在几乎不损失性能的情况下大幅减少了模型finetuning使用的显存posthttps://sgsgsama.github.io/ai/QLoRA学习笔记/AuthorSGSGPosted on2026-10-09Updated on2026-10-15Licensed under#aitrain