本地Claude Code + Gemma 4 | Local Claude Code + Gemma 4
这套组合就像你突然收养了一只“会写代码的电子猫”:不吃流量、不收月租,还能24小时帮你干活。爽是真的爽,隐私也稳得一批。唯一的小毛病是——有时候它会发呆两秒,或者突然“思考人生”,但考虑到它是白来的,这点小脾气也就显得格外可爱了。
This combo feels like adopting a “coding pet” that never eats, never bills you, and works 24/7. It’s private, free, and surprisingly powerful. Sure, it may pause to “think about life” sometimes—but hey, for something that costs nothing, those little quirks just make it charming.
核心步骤 / Combined Workflow Steps
1. 准备运行环境(安装 Node.js 和 Ollama) / Prepare the environment (install Node.js and Ollama)
2. 下载并运行本地模型(如 Gemma) / Download and run a local model (e.g., Gemma)
3. 启动本地模型服务 / Start the local model service
4. 获取 Claude Code 社区版本 / Get a community version of Claude Code
5. 安装并配置运行环境 / Install and configure the runtime environment
6. 将 Claude Code 后端指向本地模型 / Point Claude Code backend to the local model
7. 启动 Claude Code CLI / Launch the Claude Code CLI
8. 输入任务(如生成代码或修改项目) / Enter tasks (e.g., generate or modify code)
9. 自动执行并完成开发流程 / Automatically execute and complete development tasks
相关链接(relative links)
https://blog.google/technology/ai/gemma-4/
https://ollama.com
https://github.com/ollama/ollama
https://github.com/unslothai/unsloth
Comments
Post a Comment