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-webui
• ollama

ollama

安装

wget https://ollama.com/install.sh
sh install.sh
# 若网速比较慢，可借助洪荒之力
# proxychains wget https://ollama.com/install.sh
# proxychains sh install.sh
sudo systemctl daemon-reload
sudo systemctl enable ollama
sudo systemctl 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Environment="OLLAMA_MODELS=/home/ethan/Software/ollama/models"。效果如下：

ethan@G5000:/dev/shm$ cat /etc/systemd/system/ollama.service
[Unit]
Description=Ollama Service
After=network-online.target

[Service]
ExecStart=/usr/local/bin/ollama serve
User=ethan
Group=ethan
Restart=always
RestartSec=3
Environment="PATH=~/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/bin"
Environment="OLLAMA_MODELS=/home/ethan/Software/ollama/models"

[Install]
WantedBy=default.target

修改后，需要执行方可生效。

sudo systemctl daemon-reload
sudo systemctl 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 llama3.1
sudo ollama launch openclaw

常用模型

ethan@G5000:~$ cd
ethan@G5000:~$ ollama list 
NAME                                          ID              SIZE      MODIFIED   
qwen3.6:latest                                07d35212591f    23 GB     2 days ago    
gpt-oss:20b                                   17052f91a42e    13 GB     3 days ago    
llama3.1:latest                               46e0c10c039e    4.9 GB    3 days ago    
deepseek-v4-flash:cloud                       ea027821675c    -         3 days ago    
deepseek-r1:latest                            6995872bfe4c    5.2 GB    3 days ago    
deepseek-r1:8b                                6995872bfe4c    5.2 GB    3 days ago    
deepseek-coder:6.7b                           ce298d984115    3.8 GB    3 days ago    
deepseek-coder-v2:16b-lite-instruct-q4_K_M    dac6ff6589c9    10 GB     8 days ago    
qwen3-coder:latest                            06c1097efce0    18 GB     8 days ago

vscode配置

实测发现vscode默认的chat调用本地模型无法修改代码，暂不清楚原因。通过安装continue插件可以。但是吐字有点慢。通过top，intel_gpu_top和nvidia-smi dmon可发现资源都是高负荷。

安装continue插件

通过vscode的插件应用安装即可。配置文件保存在~/.continue/config.yaml。

name: Ollama Agent
version: 1.0.0
schema: v1
models:
  - name: Qwen3 Coder (Agent)
    provider: ollama
    model: qwen3-coder:latest
    apiBase: http://localhost:11434
    roles:
      - chat
      - edit
      - apply
      - summarize
      - autocomplete
      - embed
    capabilities:
      - tool_use
      - image_input
  - name: DeepSeek Coder
    provider: ollama
    model: deepseek-coder-v2:16b-lite-instruct-q4_K_M
  - name: deepseek r1
    provider: ollama
    model: deepseek-r1:8b
  - name: llama3.1
    provider: ollama
    model: llama3.1
  - name: qwen3.6:latest
    provider: ollama
    model: qwen3.6:latest
context:
  - provider: code
  - provider: docs
  - provider: diff
  - provider: terminal
  - provider: problems
  - provider: folder
  - provider: codebase
prompts: []