https://github.com/re-ovo/llm-perf
评估不同LLM提供商的性能
https://github.com/re-ovo/llm-perf
chatgpt-api deepseek-api llm
Last synced: 11 months ago
JSON representation
评估不同LLM提供商的性能
- Host: GitHub
- URL: https://github.com/re-ovo/llm-perf
- Owner: re-ovo
- License: mit
- Created: 2025-03-10T09:00:54.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-03-10T13:02:44.000Z (over 1 year ago)
- Last Synced: 2025-07-19T12:05:14.137Z (11 months ago)
- Topics: chatgpt-api, deepseek-api, llm
- Language: Python
- Homepage:
- Size: 143 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# LLM性能评估工具
> For the English version of this README, please refer to [README_EN.md](./README_EN.md).
这个工具用于评估不同LLM提供商的性能，包括以下指标：
- TTFT (Time To First Token): 从发送请求到收到第一个token的时间
- TPS (Tokens Per Second): 每秒处理的token数量
## 截图
## 配置
在`config.yml`中配置LLM提供商信息, 参考`config.example.yml`
## 安装依赖
使用[UV](https://docs.astral.sh/uv/)安装依赖
```bash
uv sync
```
## 运行
```bash
uv run main.py
```