Improve model card and surface related exports

README.md

@@ -10,24 +10,34 @@ tags:
 
 # YOLOv8-RKNN2-Ultralytics
 
-Converted with the Ultralytics export workflow and task-specific RKNN conversion tools. This repository contains 100 FP16 and INT8 exports for RK3576 and RK3588. Tasks include classify, detect, obb, pose, segment.
+> **Looking for the Rockchip Model Zoo and custom exports?** See [YOLOv8-RKNN2](https://huggingface.co/HanzoHuang/YOLOv8-RKNN2).
 
-Choose a model under `models/<toolkit-version>/<task>/<variant>/<chip>/<precision>/` and download both `model.rknn` and `metadata.yaml`. The [manifest](manifest.json) lists the supported combinations and SHA-256 hashes.
+100 FP16 and INT8 YOLOv8 exports for **RK3576** and **RK3588**, produced with the Ultralytics export workflow and task-specific RKNN conversion tools. Supported tasks include classification, detection, oriented bounding boxes, pose, and instance segmentation.
 
-## Source and license
+## Choose a model
+
+Models are organized by toolkit version, task, variant, target chip, and precision:
+
+```text
+models/v<toolkit-version>/<task>/<variant>/<chip>/<precision>/
+```
 
-The inputs are official Ultralytics checkpoints. Conversion changed the file format and quantized the INT8 graphs. The RKNN files and original weights are distributed under AGPL-3.0. See the [license](LICENSE) and [checkpoint URLs and hashes](source/weights.json).
+Download `model.rknn` and `metadata.yaml` from the same directory. The [manifest](manifest.json) lists the available models and checksums.
 
-Run `python source/convert-model.py --help`. For a detection INT8 export, pass a COCO128 dataset YAML. For other INT8 tasks, pass a COCO128 image list. Use `--calibration-limit 128` and the input size in the manifest.
+## INT8 detection
 
-The [source directory](tree/main/source) contains the original checkpoints, conversion scripts, and environment and calibration records. Upstream Ultralytics source is available from [PyPI](https://pypi.org/project/ultralytics/#history). The conversion used RKNN Toolkit2 2.3.2. See [RKNN Toolkit2](https://github.com/airockchip/rknn-toolkit2), available separately from Rockchip.
+Use the `metadata.yaml` from the same model directory. INT8 detection outputs normalized box coordinates and requires `coordinates: normalized`. Import both files together in the warehouse, or keep them in the same directory for the CLI. Treating normalized coordinates as pixels can result in no boxes.
 
-### INT8 detection
+## Reproduce the conversion
 
-Use the `metadata.yaml` from the same INT8 model folder. The detector outputs normalized box coordinates, so the app needs `coordinates: normalized`. Import both files together in the warehouse. For the CLI, keep them in the same directory. Interpreting the coordinates as pixels can result in no boxes.
+The [source directory](tree/main/source) contains the original checkpoints, conversion scripts, and environment and calibration records. Conversion used [RKNN Toolkit2 2.3.2](https://github.com/airockchip/rknn-toolkit2); upstream Ultralytics source is available on [PyPI](https://pypi.org/project/ultralytics/#history).
 
-Models are organized by RKNN Toolkit2 version under `models/v<version>/`. Each model and its metadata share a folder.
+```bash
+python source/convert-model.py --help
+```
 
-## Related RKNN exports
+For INT8 detection, pass a COCO128 dataset YAML. For other INT8 tasks, pass a COCO128 image list. Use `--calibration-limit 128` and the input size listed in the manifest.
+
+## Source and license
 
-This repository uses the Ultralytics export workflow. For the Rockchip Model Zoo conversion, see [YOLOv8-RKNN2](https://huggingface.co/HanzoHuang/YOLOv8-RKNN2).
+The inputs are official Ultralytics checkpoints. Conversion changes the format and quantizes the INT8 graphs. The RKNN files and original weights are distributed under AGPL-3.0; see [LICENSE](LICENSE) and the [checkpoint URLs and hashes](source/weights.json).