Build dataset for YOLOv4

Question

I'm learning about object detection with yolov4. I have a question about building the dataset for my data. About the output of model, the output in training mode are: [batch_size, (4 + 1 + num_classes) * num_anchors, anchor_size_1, anchor_size_1], [batch_size, (4 + 1 + num_classes) * num_anchors, anchor_size_2, anchor_size_2], [batch_size, (4 + 1 + num_classes) * num_anchors, anchor_size_3, anchor_size_3]. So to build dataset for training, what does the label matrix look like?. Ex: I have an image with 3 bounding boxes: 1 10 20 100 200, 3 5 15 150 200, 1 50 60 300 400

Tks for yours help!!