参考连接：https://modelscope.cn/models/modelscope/ollama-linux

操作系统

WEL@WEL-PC:~$ cat /etc/os-release 
PRETTY_NAME="Deepin 25"
NAME="Deepin"
VERSION_CODENAME=beige
ID=deepin
HOME_URL="https://www.deepin.org/"
BUG_REPORT_URL="https://bbs.deepin.org"
VERSION_ID="25"
VERSION="25"
WEL@WEL-PC:~$ 

操作步骤

1、安装python3.12-venv

WEL@WEL-PC:~/workspace$ sudo apt install python3.12-venv

2、创建并激活虚拟环境

WEL@WEL-PC:~/workspace$ python3 -m venv aienv
WEL@WEL-PC:~/workspace$ source aienv/bin/activate

3、设置pip清华源

(aienv) WEL@WEL-PC:~/workspace$ pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple

4、安装modelscope

(aienv) WEL@WEL-PC:~/workspace$ pip install modelscope

5、升级setuptool，解决下载时ModuleNotFoundError: No module named 'pkg_resources'错误

(aienv) WEL@WEL-PC:~/workspace$ pip install --upgrade setuptools     

错误详情：

(aienv) WEL@WEL-PC:~/workspace$ modelscope download --model=modelscope/ollama-linux --local_dir ./ollama-linux --revision v0.5.13
Traceback (most recent call last):
  File "/home/WEL/workspace/aienv/bin/modelscope", line 5, in <module>
    from modelscope.cli.cli import run_cmd
  File "/home/WEL/workspace/aienv/lib/python3.12/site-packages/modelscope/cli/cli.py", line 12, in <module>
    from modelscope.cli.plugins import PluginsCMD
  File "/home/WEL/workspace/aienv/lib/python3.12/site-packages/modelscope/cli/plugins.py", line 6, in <module>
    from modelscope.utils.plugins import PluginsManager
  File "/home/WEL/workspace/aienv/lib/python3.12/site-packages/modelscope/utils/plugins.py", line 18, in <module>
    import pkg_resources
ModuleNotFoundError: No module named 'pkg_resources'
(aienv) WEL@WEL-PC:~/workspace$ pip install pkg_resources

6、执行ollam下载

(aienv) WEL@WEL-PC:~/workspace$ modelscope download --model=modelscope/ollama-linux --local_dir ./ollama-linux --revision v0.5.13

7、切换到root用户执行安装

(aienv) WEL@WEL-PC:~/workspace/ollama-linux$ sudo su
root@WEL-PC:/home/WEL/workspace/ollama-linux# bash ollama-modelscope-install.sh

tips：如果不用root用户，安装会提示i权限不足操作失败，即使使用sudo也会失败，失败详情如下：

(aienv) WEL@WEL-PC:~/workspace/ollama-linux$ sudo bash ollama-modelscope-install.sh 
请输入密码:
验证成功
>>> Cleaning up old version at /usr/local/lib/ollama
>>> Installing Ollama to /usr/local/bin
>>> Unzipping Ollama-bundle to /usr/local...
>>> Creating ollama user...
useradd：无法创建目录 /usr/share/ollama
>>> The Ollama API is now available at 127.0.0.1:11434.
>>> Install complete. Run "ollama" from the command line.

服务验证及问题处理

1、启动ollama服务并设置开机自启

root@WEL-PC:/home/WEL/workspace# systemctl stop ollama
root@WEL-PC:/home/WEL/workspace# systemctl start ollama
root@WEL-PC:/home/WEL/workspace# systemctl status ollama
● ollama.service - Ollama Service
     Loaded: loaded (/etc/systemd/system/ollama.service; enabled; preset: enabled)
     Active: activating (auto-restart) (Result: exit-code) since Tue 2025-03-11 15:53:26 CST; 1s ago
    Process: 88906 ExecStart=/usr/local/bin/ollama serve (code=exited, status=1/FAILURE)
   Main PID: 88906 (code=exited, status=1/FAILURE)
        CPU: 43ms

查看错误信息：

WEL@WEL-PC:/persistent/home/WEL/workspace/ollama-linux$ sudo journalctl -u ollama

tips：deepin操作系统/usr/share/为系统只读，无法创建ollama

问题处理方案：ollama用户创建失败，使用当前已有用户及组

sudo vim /etc/systemd/system/ollama.service

重新加载配置及启动服务：

WEL@WEL-PC:~/.ollama$ sudo systemctl daemon-reload
WEL@WEL-PC:~/.ollama$ sudo systemctl start ollama
WEL@WEL-PC:~/.ollama$ sudo systemctl status ollama
● ollama.service - Ollama Service
     Loaded: loaded (/etc/systemd/system/ollama.service; enabled; preset: enabled)
     Active: active (running) since Tue 2025-03-11 17:28:24 CST; 17s ago
   Main PID: 116981 (ollama)
      Tasks: 12 (limit: 9168)
     Memory: 11.0M (peak: 12.1M)
        CPU: 90ms
     CGroup: /system.slice/ollama.service
             └─116981 /usr/local/bin/ollama serve

3月 11 17:28:24 WEL-PC systemd[1]: Started ollama.service - Ollama Service.
3月 11 17:28:24 WEL-PC ollama[116981]: 2025/03/11 17:28:24 routes.go:1215: INFO server config env="map[CUDA_V>
3月 11 17:28:24 WEL-PC ollama[116981]: time=2025-03-11T17:28:24.639+08:00 level=INFO source=images.go:432 msg>
3月 11 17:28:24 WEL-PC ollama[116981]: time=2025-03-11T17:28:24.640+08:00 level=INFO source=images.go:439 msg>
3月 11 17:28:24 WEL-PC ollama[116981]: time=2025-03-11T17:28:24.640+08:00 level=INFO source=routes.go:1277 ms>
3月 11 17:28:24 WEL-PC ollama[116981]: time=2025-03-11T17:28:24.640+08:00 level=INFO source=gpu.go:217 msg="l>
3月 11 17:28:24 WEL-PC ollama[116981]: time=2025-03-11T17:28:24.677+08:00 level=WARN source=amd_linux.go:61 m>
3月 11 17:28:24 WEL-PC ollama[116981]: time=2025-03-11T17:28:24.678+08:00 level=INFO source=amd_linux.go:402 >
3月 11 17:28:24 WEL-PC ollama[116981]: time=2025-03-11T17:28:24.678+08:00 level=INFO source=gpu.go:377 msg="n>
3月 11 17:28:24 WEL-PC ollama[116981]: time=2025-03-11T17:28:24.678+08:00 level=INFO source=types.

2、拉取模型：WEL@WEL-PC:~/.ollama$ ollama run deepseek-r1:1.5b