search

chaofengc / PSFRGAN

PyTorch codes for "Progressive Semantic-Aware Style Transformation for Blind Face Restoration", CVPR2021
Other
370 stars 68 forks source link

CUDA out of memory. #25

Closed monsterlyg closed 3 years ago

monsterlyg commented 3 years ago

are you sure that the model can train on a single GPU with 8G memory.? As it needs to load four(G, D, P, VGG) networks during training. I cannot run train with a 12-G gpu.

chaofengc commented 3 years ago

Please check your batch size. I trained the model with batch_size=2 on two 1080Ti GPUs.

BTW, I do not have GPU device larger than 12-G gpu to train the model.