tyloocifer

i used co_dino_5scale_lsj_r50_1x_coco.py in the MMdetection project

When i change the model to dino it works. but co_dino doesnt. i try to reduce the num_co_head and only use fasterRcnn or Atss it still require a lot of...

When i change the model to dino it works. but co_dino doesnt. i try to reduce the num_co_head and only use fasterRcnn or Atss it still require a lot of...

it still doesn`t work. I just adopt your network. As for dataloader and other config i didn`t use it. when I use Dino it just allocate 10G when batchsize=1, but...

yep

perhaps i need to change it into 4scale?

if i wanna get a 4-scale model, where should i change except config file.

The total loss has been oscillating around 20, is this normal?