93 lines
2.2 KiB
Markdown
93 lines
2.2 KiB
Markdown
---
|
||
title: "在 Ubuntu 安装 Ollama 并运行 Qwen2.5-Coder 7B"
|
||
type: source
|
||
tags: [ollama, qwen, qwen-coder, ubuntu, local-llm]
|
||
date: 2026-04-13
|
||
source_file: raw/Technical/Home Office/在 Ubuntu 安装 Ollama 并运行 Qwen2.5‑Coder 7B.md
|
||
---
|
||
|
||
## Summary
|
||
- 核心主题：Ubuntu服务器安装Ollama运行Qwen2.5-Coder 7B本地大模型
|
||
- 问题域：本地LLM部署、模型服务化、API开放
|
||
- 方法/机制：Ollama一键安装 + systemd服务 + REST API暴露
|
||
- 结论/价值：构建本地AI编码助手基础设施
|
||
|
||
## Key Claims
|
||
- Ollama提供一键式LLM部署
|
||
- Qwen2.5-Coder 7B适合代码生成和DevOps任务
|
||
- 默认只监听127.0.0.1，需配置才能远程访问
|
||
- GPU加速开箱即用（需CUDA）
|
||
|
||
## Key Concepts
|
||
- [[本地LLM]]：本地部署的大语言模型
|
||
- [[Ollama]]：LLM运行平台
|
||
- [[Qwen2.5-Coder]]：阿里通义千问代码模型
|
||
- [[REST API]]：模型服务接口
|
||
|
||
## Key Entities
|
||
- [[Ollama]]：LLM服务平台
|
||
- [[Qwen2.5-Coder]]：代码生成模型
|
||
- [[Ubuntu]]：部署系统
|
||
|
||
## 系统要求
|
||
| 资源 | 最低 | 推荐 |
|
||
|------|------|------|
|
||
| CPU | 4核 | 8+核 |
|
||
| RAM | 8GB | 16GB |
|
||
| GPU | 无需 | NVIDIA |
|
||
| 磁盘 | 10GB | 20GB |
|
||
|
||
模型大小：约4.5GB
|
||
|
||
## 安装步骤
|
||
```bash
|
||
# 1. 安装
|
||
curl -fsSL https://ollama.com/install.sh | sh
|
||
|
||
# 2. 下载模型
|
||
ollama pull qwen2.5-coder:7b
|
||
|
||
# 3. 运行
|
||
ollama run qwen2.5-coder:7b
|
||
```
|
||
|
||
## API调用
|
||
|
||
### REST API
|
||
```bash
|
||
curl http://localhost:11434/api/chat -d '{
|
||
"model": "qwen2.5-coder:7b",
|
||
"messages": [{"role": "user", "content": "Write a bash script"}]
|
||
}'
|
||
```
|
||
|
||
### Python调用
|
||
```python
|
||
from ollama import chat
|
||
response = chat(model="qwen2.5-coder:7b", messages=[{"role": "user", "content": "..."}])
|
||
```
|
||
|
||
### NodeJS调用
|
||
```javascript
|
||
import ollama from 'ollama'
|
||
const response = await ollama.chat({ model: 'qwen2.5-coder:7b', messages: [...] })
|
||
```
|
||
|
||
## 开放远程API
|
||
编辑`/etc/systemd/system/ollama.service`，添加：
|
||
```ini
|
||
Environment="OLLAMA_HOST=0.0.0.0"
|
||
```
|
||
然后：
|
||
```bash
|
||
sudo systemctl daemon-reload
|
||
sudo systemctl restart ollama
|
||
```
|
||
|
||
## 推荐搭配
|
||
| 工具 | 用途 |
|
||
|------|------|
|
||
| Open WebUI | ChatGPT UI |
|
||
| n8n | AI自动化 |
|
||
| OpenClaw | AI coding agent |
|
||
| LangChain | Agent framework | |