as the title says, you have two scripts, run_bsl_resnet38 and run _voc_resnet38
what the difference between the two?
Is its purpose to create pretrained models? so if i run bls_resnet first, the model its outputs can be used to replace the initial weights from ilsvrc-cls_rna-a1_cls1000_ep-0001.pth (402M)?