I used byol.yaml to train resnet50 backbone with BYOL and custom dataset. When I converted the trained model to detectron2 model and used it with train_object_detection.py and coco_R_50_C4_2x.yaml, I got error that GT data does not have gt masks. However, I wanted to train this and use FRCNN as my work and not MRCNN which requires mask data.
Hi,
I used byol.yaml to train resnet50 backbone with BYOL and custom dataset. When I converted the trained model to detectron2 model and used it with train_object_detection.py and coco_R_50_C4_2x.yaml, I got error that GT data does not have gt masks. However, I wanted to train this and use FRCNN as my work and not MRCNN which requires mask data.
How can I solve this error?