facebookresearch / Mask2Former

Code release for "Masked-attention Mask Transformer for Universal Image Segmentation"
MIT License
2.59k stars 388 forks source link

how use custom pre-trained backbone in mask2former #226

Open sleepat-11 opened 10 months ago

sleepat-11 commented 10 months ago

Your work is excellent and impressive! But I have a little question how to load my own pre-trained backbone to replace the origin backbone, for example .i got a vit pre-trained model ,which is a pth or pt file, how to load pth in detectron2? do i have to tranfer it into pkl file by convert-torchvision-to-d2.py? Thanks again for your great work!

sleepat-11 commented 10 months ago

and i also need to freeze my backbone parameters? how to do that?