---
license: apache-2.0
tags:
- object-detection
- yolox
- pytorch
- safetensors
library_name: yolox
pipeline_tag: object-detection
---

# yolox-tiny

This repository contains **YOLOX-tiny** weights exported to **`safetensors`** format.

## Source

These weights are derived from the official **YOLOX** project by **Megvii-BaseDetection**:

- Official repository: https://github.com/Megvii-BaseDetection/YOLOX

Original checkpoint source used for this export:

- `weights/yolox_tiny.pth`

## License

YOLOX is released under the **Apache-2.0** license.  
Please refer to the original project for the full license terms.

## Files

This repository is expected to contain:

- `model.safetensors` — model weights
- `config.json` — lightweight metadata
- `LICENSE` — copied license file
- `README.md` — this model card

## Official YOLOX Benchmark

The following reference metrics are taken from the official YOLOX benchmark table.

| Model | Input size | mAP val (0.5:0.95) | Params (M) | FLOPs (G) |
|---|---:|---:|---:|---:|
| YOLOX-Tiny | 416 | 32.8 | 5.06 | 6.45 |

## Notes

This is **not an official YOLOX release**.  
This repository republishes model weights in **`safetensors`** format for easier sharing and hosting on Hugging Face Hub.

## Acknowledgements

Thanks to the **YOLOX authors and maintainers** for the original work.