about the DINO training loss

facebookresearch / dino

PyTorch code for Vision Transformers training with the Self-Supervised learning method DINO

Apache License 2.0

6.38k stars 910 forks source link

about the DINO training loss #225

Open Faisal-Hajari opened 1 year ago

Faisal-Hajari commented 1 year ago

Hello, I'm training resnet18 on a custom dataset. it's been running for some time with a batch size of 325 (the max my gpu can handle). the thing is the loss is flat and it's not getting better or worse. is this behavior normal ? and if so how do you decide on stopping the training ?

ComeBackCity commented 1 year ago

Is it stuck around the 10-11 range?