Closed tms2003 closed 5 months ago
Our base model is a generic semantic segmentation model such as UNetfomer or a ResNet18, not Sgement Anything Model. So I'm not sure I understand the meaning of your fine-tuning. Can you describe in more detail what you want to do?
I want to fine-tune SAM_RS, can you tell me the basic steps (using LoveDA as an example, although you have already used it)?