📄 中文摘要
通过将一部旧的 4GB 安卓手机转换为本地 LLM 服务器，实现了离线运行 AI 的目标。这一实验使用了 Termux 创建类 Linux 环境，Ollama 运行本地语言模型，以及 Qwen2（0.5b 变体）作为轻量级模型。此举不仅避免了云 AI 的高昂费用和对互联网的依赖，还展示了边缘 AI 的未来潜力。实验中总结了有效的方法与遇到的挑战，为个人用户提供了新的选择。
在本地运行 AI：将安卓手机转变为个人 LLM 服务器
📄 English Summary
Run AI Locally: Converting an Android phone into into a personal LLM server
An old 4GB Android phone was successfully converted into a local LLM server, achieving the goal of running AI offline. The experiment utilized Termux to create a Linux-like environment, Ollama to run local language models, and Qwen2 (0.5b variant) as a lightweight model. This approach not only avoids the high costs associated with cloud AI and reliance on internet access but also highlights the potential of edge AI for personal users. The experiment outlines what worked, what didn't, and offers insights into future possibilities.