参考文章Visual Studio 中的 AI 辅助开发基于Ollama的本地大模型自动化编程实践指南open-webuiollamaollama安装wgethttps://ollama.com/install.shshinstall.sh# 若网速比较慢可借助洪荒之力# proxychains wget https://ollama.com/install.sh# proxychains sh install.shsudosystemctl daemon-reloadsudosystemctlenableollamasudosystemctl restart ollama如果需要自定义模型保存路径可修改服务文件增加环境变量EnvironmentOLLAMA_MODELS/home/ethan/Software/ollama/models。效果如下ethanG5000:/dev/shm$cat/etc/systemd/system/ollama.service[Unit]DescriptionOllama ServiceAfternetwork-online.target[Service]ExecStart/usr/local/bin/ollama serveUserethanGroupethanRestartalwaysRestartSec3EnvironmentPATH~/Self/repo:~/Self/python/venv/bin:~/Self/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/snap/bin:/snap/binEnvironmentOLLAMA_MODELS/home/ethan/Software/ollama/models[Install]WantedBydefault.target修改后需要执行方可生效。sudosystemctl daemon-reloadsudosystemctl restart ollama常用命令ollama list ollama show [model] ollama run [model]例如ollama list ollama show qwen3-coder:latest ollama run llama3.1sudoollama launch openclawvscode配置安装continue插件name:Ollama Agentversion:1.0.0schema:v1models:-name:Qwen3 Coder (Agent)provider:ollamamodel:qwen3-coder:latestapiBase:http://localhost:11434roles:-chat-edit-apply-summarize-autocomplete-embedcapabilities:-tool_use-image_input-name:DeepSeek Coderprovider:ollamamodel:deepseek-coder-v2:16b-lite-instruct-q4_K_M-name:deepseek r1provider:ollamamodel:deepseek-r1:8b-name:llama3.1provider:ollamamodel:llama3.1-name:qwen3.6:latestprovider:ollamamodel:qwen3.6:latestcontext:-provider:code-provider:docs-provider:diff-provider:terminal-provider:problems-provider:folder-provider:codebaseprompts:[]
