This is the official implementation of the paper Lesion-based Contrastive Learning for Diabetic Retinopathy Grading from Fundus Images.
38
stars
6
forks
source link
How many GPUs did you use for training the detection model? #9
Closed
osivaz61 closed 6 months ago
We use only one GPU with 24GB Memory. The batch size for training the detection model does not have to be large.