Excuse me, if I need to develop a second time based on GLIP and migrate its pre-trained model to my own downstream task (my own dataset), should I use fine tuning or prompt tuning? What is the corresponding command? Additionally, What is the difference between tools/train_net.py and tools/finetune.py?
Excuse me, if I need to develop a second time based on GLIP and migrate its pre-trained model to my own downstream task (my own dataset), should I use fine tuning or prompt tuning? What is the corresponding command? Additionally, What is the difference between tools/train_net.py and tools/finetune.py?