mahmoodlab / HIPT

Hierarchical Image Pyramid Transformer - CVPR 2022 (Oral)
Other
498 stars 86 forks source link

Pretrained 3. ViT? #74

Open vildesboe opened 4 months ago

vildesboe commented 4 months ago

Hi! Thank you for publishing your code and for continuing to answer questions in this repo. I am working on the survival task, and wonder if you pretrain the final ViT in any way, since you say you just "finetune" it for 20 epochs? Or is it finetuned from a random initialization? Best regards