<?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>Pipeline on 硅基 AGI · 智能体学习与测评</title><link>https://guijiagi.com/tags/pipeline/</link><description>Recent content in Pipeline 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>Thu, 25 Jun 2026 10:00:00 +0800</lastBuildDate><atom:link href="https://guijiagi.com/tags/pipeline/index.xml" rel="self" type="application/rss+xml"/><item><title>Haystack 2.0 评测：deepset 的 RAG 框架重生</title><link>https://guijiagi.com/posts/haystack-20-review/</link><pubDate>Thu, 25 Jun 2026 10:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/haystack-20-review/</guid><description>Haystack 2.0 相比 1.x 有哪些根本性变化？Pipeline 组件架构如何设计？与 LangChain 相比在企业 RAG 部署上有何优势？本文全面拆解。</description></item><item><title>LLM 微调流水线设计：从数据到部署的 MLOps</title><link>https://guijiagi.com/posts/llm-finetune-pipeline/</link><pubDate>Thu, 25 Jun 2026 10:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/llm-finetune-pipeline/</guid><description>完整的 LLM 微调 MLOps 流水线：数据准备、训练配置、评估、版本管理、A/B 测试与灰度发布</description></item><item><title>高级 Prompt 链式调用：构建复杂推理流水线</title><link>https://guijiagi.com/posts/prompt-chaining-advanced/</link><pubDate>Wed, 24 Jun 2026 16:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/prompt-chaining-advanced/</guid><description>深入探讨 Prompt 链式调用的架构模式，构建可靠的 LLM 推理流水线</description></item></channel></rss>