Spaces:
Sleeping
Sleeping
Commit
·
d2d545d
1
Parent(s):
ab07c5a
readme
Browse files
README.md
CHANGED
|
@@ -1,167 +0,0 @@
|
|
| 1 |
-
# Abacus2Api
|
| 2 |
-
|
| 3 |
-
<div align="center">
|
| 4 |
-
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
[](https://fastapi.tiangolo.com/)
|
| 8 |
-
[](https://www.python.org/)
|
| 9 |
-
[](LICENSE)
|
| 10 |
-
|
| 11 |
-
_✨ 一个API代理服务 ✨_
|
| 12 |
-
|
| 13 |
-
</div>
|
| 14 |
-
|
| 15 |
-
## 📖 项目介绍
|
| 16 |
-
|
| 17 |
-
Abacus2Api是一个基于FastAPI构建的API代理服务，它提供了一个统一的接口来与各种大型语言模型(LLM)进行流式交互。该服务完全兼容OpenAI API格式，让您可以轻松地将现有应用与不同的AI模型集成。
|
| 18 |
-
|
| 19 |
-
### 主要特性
|
| 20 |
-
|
| 21 |
-
- 🚀 兼容OpenAI API格式
|
| 22 |
-
- 🌊 原生支持流式(Stream)响应
|
| 23 |
-
- 🔄 统一接口访问多种LLM模型
|
| 24 |
-
- 🛡️ 内置错误处理和异常管理
|
| 25 |
-
- 🌐 支持CORS，方便前端集成
|
| 26 |
-
|
| 27 |
-
## 🔧 系统架构
|
| 28 |
-
|
| 29 |
-
```mermaid
|
| 30 |
-
flowchart TD
|
| 31 |
-
A[客户端应用] -->|API请求| B
|
| 32 |
-
B[Abacus2Api] -->|处理请求| C
|
| 33 |
-
C[消息格式化] --> D
|
| 34 |
-
D[流式处理] --> E
|
| 35 |
-
E[响应生成] --> F
|
| 36 |
-
F[返回客户端]
|
| 37 |
-
|
| 38 |
-
style B fill:#1e88e5,color:white
|
| 39 |
-
style D fill:#43a047,color:white
|
| 40 |
-
style F fill:#e53935,color:white
|
| 41 |
-
```
|
| 42 |
-
|
| 43 |
-
## 💻 安装指南
|
| 44 |
-
|
| 45 |
-
### 前置条件
|
| 46 |
-
|
| 47 |
-
- Python 3.8+
|
| 48 |
-
- pip (Python包管理工具)
|
| 49 |
-
|
| 50 |
-
### 安装步骤
|
| 51 |
-
|
| 52 |
-
1. 克隆仓库
|
| 53 |
-
|
| 54 |
-
```bash
|
| 55 |
-
git clone https://github.com/yourusername/abacus2api.git
|
| 56 |
-
cd abacus2api
|
| 57 |
-
```
|
| 58 |
-
|
| 59 |
-
2. 创建并激活虚拟环境（可选但推荐）
|
| 60 |
-
|
| 61 |
-
```bash
|
| 62 |
-
# Windows
|
| 63 |
-
python -m venv venv
|
| 64 |
-
venv\Scripts\activate
|
| 65 |
-
|
| 66 |
-
# macOS/Linux
|
| 67 |
-
python -m venv venv
|
| 68 |
-
source venv/bin/activate
|
| 69 |
-
```
|
| 70 |
-
|
| 71 |
-
3. 安装依赖
|
| 72 |
-
|
| 73 |
-
```bash
|
| 74 |
-
pip install -r requirements.txt
|
| 75 |
-
```
|
| 76 |
-
|
| 77 |
-
## 🚀 使用方法
|
| 78 |
-
|
| 79 |
-
### 启动服务
|
| 80 |
-
|
| 81 |
-
```bash
|
| 82 |
-
uvicorn main:app --reload --host 0.0.0.0 --port 8000
|
| 83 |
-
```
|
| 84 |
-
|
| 85 |
-
服务将在 `http://localhost:8000` 上运行。
|
| 86 |
-
|
| 87 |
-
### API端点
|
| 88 |
-
|
| 89 |
-
#### 健康检查
|
| 90 |
-
|
| 91 |
-
```
|
| 92 |
-
GET /
|
| 93 |
-
```
|
| 94 |
-
|
| 95 |
-
#### 获取可用模型列表
|
| 96 |
-
|
| 97 |
-
```
|
| 98 |
-
GET /v1/models
|
| 99 |
-
```
|
| 100 |
-
|
| 101 |
-
#### 创建聊天完成（流式响应）
|
| 102 |
-
|
| 103 |
-
```
|
| 104 |
-
POST /v1/chat/completions
|
| 105 |
-
```
|
| 106 |
-
|
| 107 |
-
请求体示例:
|
| 108 |
-
|
| 109 |
-
```json
|
| 110 |
-
{
|
| 111 |
-
"messages": [
|
| 112 |
-
{
|
| 113 |
-
"role": "system",
|
| 114 |
-
"content": "你是一个有用的AI助手。"
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"role": "user",
|
| 118 |
-
"content": "你好，请介绍一下自己。"
|
| 119 |
-
}
|
| 120 |
-
],
|
| 121 |
-
"model": "gpt-3.5-turbo",
|
| 122 |
-
"stream": true
|
| 123 |
-
}
|
| 124 |
-
```
|
| 125 |
-
|
| 126 |
-
## 📊 请求流程
|
| 127 |
-
|
| 128 |
-
```mermaid
|
| 129 |
-
sequenceDiagram
|
| 130 |
-
participant C as 客户端
|
| 131 |
-
participant A as Abacus2Api
|
| 132 |
-
participant L as 语言模型
|
| 133 |
-
|
| 134 |
-
C->>A: 发送请求
|
| 135 |
-
A->>A: 处理请求
|
| 136 |
-
A->>L: 转发请求
|
| 137 |
-
L-->>A: 流式响应
|
| 138 |
-
loop 响应块
|
| 139 |
-
A-->>C: 发送事件
|
| 140 |
-
end
|
| 141 |
-
```
|
| 142 |
-
|
| 143 |
-
## 🔍 故障排除
|
| 144 |
-
|
| 145 |
-
常见问题:
|
| 146 |
-
|
| 147 |
-
1. **连接超时**
|
| 148 |
-
- 检查网络连接
|
| 149 |
-
- 确认LLM服务端点是否可访问
|
| 150 |
-
|
| 151 |
-
2. **认证错误**
|
| 152 |
-
- 验证API密钥是否有效
|
| 153 |
-
- 确保Authorization格式正确
|
| 154 |
-
|
| 155 |
-
3. **请求格式错误**
|
| 156 |
-
- 遵循API文档中的请求格式
|
| 157 |
-
- 检查必填字段
|
| 158 |
-
|
| 159 |
-
## 📄 许可证
|
| 160 |
-
|
| 161 |
-
本项目基于MIT许可证发布 - 详情请查看[LICENSE](LICENSE)文件。
|
| 162 |
-
|
| 163 |
-
---
|
| 164 |
-
|
| 165 |
-
<div align="center">
|
| 166 |
-
Made with ❤️ by Abacus Team
|
| 167 |
-
</div>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|