<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>CPT on 硅基 AGI · 智能体学习与测评</title><link>https://guijiagi.com/tags/cpt/</link><description>Recent content in CPT on 硅基 AGI · 智能体学习与测评</description><generator>Hugo</generator><language>zh-cn</language><copyright>本站内容采用 CC BY-NC-SA 4.0 国际许可协议授权</copyright><lastBuildDate>Sun, 28 Jun 2026 10:00:00 +0800</lastBuildDate><atom:link href="https://guijiagi.com/tags/cpt/index.xml" rel="self" type="application/rss+xml"/><item><title>持续预训练实践：让开源模型学会领域知识</title><link>https://guijiagi.com/posts/continual-pretraining-domain-knowledge-injection/</link><pubDate>Sun, 28 Jun 2026 10:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/continual-pretraining-domain-knowledge-injection/</guid><description>深入讲解大模型持续预训练（CPT）的工程实践，涵盖数据配比、训练策略、灾难性遗忘的预防和评估方法</description></item><item><title>持续预训练实践：领域大模型的训练方法论</title><link>https://guijiagi.com/posts/continuous-pretraining/</link><pubDate>Thu, 25 Jun 2026 12:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/continuous-pretraining/</guid><description>系统讲解大模型持续预训练（CPT）的完整方法论：领域数据构建、训练策略、灾难性遗忘缓解、学习率调度、数据混合比例、评估体系，附完整训练代码与企业级实践指南。</description></item></channel></rss>