https://thinkless-github-io.pages.dev/tags/%E6%8E%A8%E7%90%86%E9%83%A8%E7%BD%B2/Recent content in 推理部署 on 杨の草原Hugozh-CNFri, 19 Dec 2025 17:14:09 +0800https://thinkless-github-io.pages.dev/posts/%E5%A4%9A%E6%A8%A1%E6%80%81%E6%8E%A8%E7%90%86%E6%A8%A1%E5%9E%8B/Fri, 19 Dec 2025 17:14:09 +0800https://thinkless-github-io.pages.dev/posts/%E5%A4%9A%E6%A8%A1%E6%80%81%E6%8E%A8%E7%90%86%E6%A8%A1%E5%9E%8B/梳理大型多模态推理模型（LMRMs）的技术演进路线：从早期感知驱动的模块化设计，到大模型时代的思维链（CoT）推理，再到基于强化学习的长程规划系统。https://thinkless-github-io.pages.dev/posts/vllm%E6%8E%A8%E7%90%86%E6%80%A7%E8%83%BD%E5%8E%8B%E6%B5%8B/Thu, 17 Jul 2025 21:02:19 +0800https://thinkless-github-io.pages.dev/posts/vllm%E6%8E%A8%E7%90%86%E6%80%A7%E8%83%BD%E5%8E%8B%E6%B5%8B/vLLM 推理性能压测笔记，记录 PagedAttention 内存分配、KV Cache 优化、参数调优和吞吐量测试结果。https://thinkless-github-io.pages.dev/posts/vllm/Wed, 21 May 2025 21:35:30 +0800https://thinkless-github-io.pages.dev/posts/vllm/vLLM 推理框架笔记，记录 PagedAttention 内存管理、量化技术、分布式部署和 OpenAI 兼容 API 的使用方式。https://thinkless-github-io.pages.dev/posts/%E6%8E%A8%E7%90%86%E6%A1%86%E6%9E%B6/Tue, 20 May 2025 10:20:29 +0800https://thinkless-github-io.pages.dev/posts/%E6%8E%A8%E7%90%86%E6%A1%86%E6%9E%B6/我自己的推理框架实战笔记：ONNX、TensorRT、TorchScript，聊原理也给出部署踩坑与优化经验。https://thinkless-github-io.pages.dev/posts/ollama%E9%83%A8%E7%BD%B2/Fri, 09 May 2025 11:08:46 +0800https://thinkless-github-io.pages.dev/posts/ollama%E9%83%A8%E7%BD%B2/记录 Ollama 本地部署流程，包括安装配置、模型下载、GGUF 格式导入、自定义 Modelfile 创建和运行管理。https://thinkless-github-io.pages.dev/posts/openweb-ui%E6%8C%87%E5%8D%97%E5%9F%BA%E4%BA%8Edocker%E5%AE%89%E8%A3%85/Tue, 29 Apr 2025 21:29:21 +0800https://thinkless-github-io.pages.dev/posts/openweb-ui%E6%8C%87%E5%8D%97%E5%9F%BA%E4%BA%8Edocker%E5%AE%89%E8%A3%85/OpenWeb UI Docker 部署指南，记录 Windows WSL2 环境配置、网络代理设置和镜像源优化。解决常见安装问题，快速搭建本地 AI 聊天界面。