Metadata-Version: 2.4
Name: ahu-paimon-toolkit
Version: 0.1.0
Summary: A special toolkit for SiriusAhu's FYP, Player-Aware Intelligent Monitoring and Operations Navigator
Author-email: SiriusAhu <SiriusLLLUUU@gmail.com>
Requires-Python: >=3.12
Requires-Dist: loguru>=0.7.3
Requires-Dist: nvidia-ml-py>=13.590.48
Requires-Dist: openai>=2.24.0
Requires-Dist: ruff>=0.15.4
Description-Content-Type: text/markdown

# Ahu's Paimon Toolkit

This is a toolkit for SiriusAhu's FYP, *PAIMON* (Player-Aware Intelligent Monitoring and Operations Navigator).

## `vLLM` Benchmark

This toolkit provides a series of benchmark functions for `vLLM` server.

**Note**: The `vLLM` benchmark needs the download of `QWEN3-VL-2B-Instruct` model, which could be automatically downloaded by running benchmark functions.

To speed up the download process, China users can set environment variable `HF_ENDPOINT` to `https://hf-mirror.com` before running benchmark functions.

```bash
export HF_ENDPOINT=https://hf-mirror.com
```

## Usage

```bash
python -m ahu_paimon_toolkit.benchmark_vllm
```

## License