When I set the Trainer as baseline in the faster_rcnn_VGG_crosscity.yaml, I get the following error:
FileNotFoundError: [Errno 2] No such file or directory: 'dataseed/COCO_supervision.txt'
this is even when the DATALOADER.SUP_PERCENT = 100
I didn't understand what is the importance of the dataseed/COCO_supervision.txt file
How is limiting the percentage of labelled and unlabelled dataset helping the model learn?
When I set the Trainer as baseline in the faster_rcnn_VGG_crosscity.yaml, I get the following error:
FileNotFoundError: [Errno 2] No such file or directory: 'dataseed/COCO_supervision.txt' this is even when the DATALOADER.SUP_PERCENT = 100
I didn't understand what is the importance of the dataseed/COCO_supervision.txt file How is limiting the percentage of labelled and unlabelled dataset helping the model learn?