Dataset and annotations for pretraining YOLO-World-Seg

[takakatsuki]

I would like to fine-tune YOLO World for segmentation. However, I cannot find how to prepare a dataset for fine-tuning in your documents (https://github.com/AILab-CVC/YOLO-World/blob/master/docs/data.md). Could you give me a pointer to the datasets you are using pre-training, such as' lvis_v1_train_base.json`?

[wondervictor]

Hi @takakatsuki, if you only want to fine-tune YOLO-World-Seg, you can change it to lvis_v1_train.json.