---
id: 20260819-T0-01
title: "新训练方法FPO让模型微调提速3倍，省40%内存"
title_en: "Forward-Only Training Boosts LLM Fine-tuning 3x, Cuts Memory 40%"
url: https://ai.daily.yangsir.net/daily/20260819-T0-01
issue_date: 2026-08-19
publish_date: 2026-08-18T04:00:00.000Z
category: research
source_name: "arXiv cs.LG (ML)"
source_url: https://arxiv.org/abs/2608.14563
---

# 新训练方法FPO让模型微调提速3倍，省40%内存

arXiv新论文提出前向传播域适应（FPO）方法，训练大语言模型时无需通过模型主体进行反向传播。相比标准微调，吞吐量提升2.7至3.2倍，峰值训练内存降低约40%，同时域外基准测试性能保持稳定。这项技术有望大幅降低大模型微调的门槛，让资源有限的团队也能高效适配模型。

## English Version

**Forward-Only Training Boosts LLM Fine-tuning 3x, Cuts Memory 40%**

A new arXiv paper introduces Forward-Pass-Only (FPO) domain adaptation, a method that trains LLMs without backpropagation through the model body. It achieves 2.7-3.2x the throughput of standard fine-tuning while using roughly 40% less peak memory, with off-domain benchmarks remaining stable. This could enable efficient model adaptation for resource-constrained teams.

---

**来源**：[arXiv cs.LG (ML)](https://arxiv.org/abs/2608.14563)

**详情页**：https://ai.daily.yangsir.net/daily/20260819-T0-01

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*