<?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>本地部署 on 技术博客 - 有条工具</title><link>/blog/tags/%E6%9C%AC%E5%9C%B0%E9%83%A8%E7%BD%B2/</link><description>Recent content in 本地部署 on 技术博客 - 有条工具</description><generator>Hugo -- 0.147.7</generator><language>zh-cn</language><copyright>2024-2025 有条工具技术博客</copyright><lastBuildDate>Sat, 03 Oct 2026 10:00:00 +0800</lastBuildDate><atom:link href="/blog/tags/%E6%9C%AC%E5%9C%B0%E9%83%A8%E7%BD%B2/feed.xml" rel="self" type="application/rss+xml"/><item><title>GGUF 量化怎么选：Q3 / Q4 / Q6 / Q8 的质量与体积取舍</title><link>/blog/articles/gguf-quantization-guide/</link><pubDate>Sat, 03 Oct 2026 10:00:00 +0800</pubDate><guid>/blog/articles/gguf-quantization-guide/</guid><description>量化是大模型进消费级硬件的门票。解读 GGUF 文件命名规则，给出各档位体积与质量对照表，并按显存富余程度给出选择决策树。</description></item><item><title>本地部署还是调 API？一笔算得清的成本账</title><link>/blog/articles/local-vs-api-cost/</link><pubDate>Sat, 03 Oct 2026 09:15:00 +0800</pubDate><guid>/blog/articles/local-vs-api-cost/</guid><description>API 按 Token 计费，本地只花电费——但折旧、并发与质量差距也要算进去。用一套可复算的成本模型，给轻度、中度、重度三档调用量的明确结论。</description></item><item><title>Ollama 从安装到跑通：新手完整教程与常见报错</title><link>/blog/articles/ollama-getting-started-guide/</link><pubDate>Sat, 03 Oct 2026 09:05:00 +0800</pubDate><guid>/blog/articles/ollama-getting-started-guide/</guid><description>手把手安装 Ollama、拉取模型、调通 REST API，覆盖 OOM、生成速度慢、连接拒绝等高频报错的排查思路，新手照做即可跑通。</description></item></channel></rss>