# Third-party notice — YOLOX-Nano

This directory contains an unmodified pretrained model checkpoint redistributed under its original license.

- **Model**: YOLOX-Nano (object detection, COCO 80 classes)
- **Upstream repository**: https://github.com/Megvii-BaseDetection/YOLOX
- **Copyright**: Copyright (c) Megvii, Inc. and its affiliates
- **License**: Apache License 2.0 (see `LICENSE` in this directory, copied verbatim from the upstream repository)
- **Commercial use**: Permitted. Apache-2.0 has no non-commercial restriction; it only requires retaining the copyright/license notice and stating any changes made to the licensed material.
- **Modifications made**: None. `yolox_nano.onnx` is used exactly as published upstream; no weights were retrained or altered.
- **Asset source**: https://github.com/Megvii-BaseDetection/YOLOX/releases/download/0.1.1rc0/yolox_nano.onnx
- **Release tag**: `0.1.1rc0`
- **File size**: 3,659,407 bytes
- **SHA-256**: `c789161ed43c8269fcd4e67c67eeeb4e80c622da2eb296a20bc6007bd18a0b7d`
- **Training dataset**: COCO (Common Objects in Context) — annotations licensed CC BY 4.0; no non-commercial restriction on the resulting model weights.

Preprocessing/postprocessing logic used in this site's `demo.js` was independently implemented in JavaScript from YOLOX's published `data_augment.py` / `demo_utils.py` reference logic (letterbox resize, grid/stride decode, class-agnostic NMS) and is original code, not redistributed from the upstream repository.