Feature Request: Splitting Dataset that is in Yolo .txt format

[richard-schwab]

Hello,

So in reading https://github.com/obss/sahi/blob/main/docs/slicing.md I noticed the current workflow for yolo .txt datasets is:

1. Convert the yolo annotations to coco format.
2. Slice the coco annotations using sahi.
3. Convert sliced coco files into yolo format using sahi
4. Train yolo with the converted yolo txt files

Is there any way we could slice yolo formatted files as well? Perhaps add a slice_yolo_annotations function to https://github.com/obss/sahi/blob/main/sahi/slicing.py?

Ideally this would be the workflow:

1. Slice a yolo annotation using sahi.
2. Train yolo with the sliced yolo txt files.