yolov8 docs update (#69)

* yolov8 docs update

site/content/en/docs/formats/yolo_ultralytics.md

@@ -179,6 +179,22 @@ All coordinates must be normalized and be in range \[0, 1\].
 It can be achieved by dividing x coordinates and widths by image width,
 and y coordinates and heights by image height.
 
+For compatibility with other tools exporting in Ultralytics YOLO format
+(e.g. [roboflow](https://roboflow.com/formats/yolov8-pytorch-txt)),
+datumaro supports datasets with the inverted directory order of subset and "images" or "labels",
+i.e. both `train/images/`, `images/train/` are valid inputs.
+```bash
+archive.zip/
+   ├── train/
+   │   ├── images/  # directory with images for train subset
+   │   │    ├── image1.jpg
+   │   │    ├── image2.jpg
+   │   │    └── ...
+   │   ├── labels/  # directory with annotations for train subset
+   │   │    ├── image1.txt
+   │   │    ├── image2.txt
+   │   │    └── ...
+```
 
 ## Export to other formats