search

zudi-lin / rcan-it

Revisiting RCAN: Improved Training for Image Super-Resolution
MIT License
91 stars 16 forks source link

how to fine-tune model with large-patch #4

Closed chenjiachengzzz closed 2 years ago

chenjiachengzzz commented 2 years ago

hello, could you give me a example for fine-tune model with large-patch..

zudi-lin commented 2 years ago

Hi @chenjiachengzzz, below is the Linux command we use for large-patch finetuning, with 4 GPUs

CUDA_VISIBLE_DEVICES=0,1,2,3 python -u -m torch.distributed.run --nproc_per_node=4 \
--master_port=8889 main.py --distributed --config-base configs/RCAN/RCAN_Improved.yaml \
--config-file configs/RCAN/RCAN_x2_LP.yaml SOLVER.ITERATION_RESTART True SOLVER.ITERATION_TOTAL 40000 \
SOLVER.BASE_LR 8e-4 MODEL.PRE_TRAIN baseline/RCAN_x2_baseline/model/model_latest.pth.tar

Please change the checkpoint to your pretrained model path. Let me know if you have additional questions!

chenjiachengzzz commented 2 years ago

Hi @chenjiachengzzz, below is the Linux command we use for large-patch finetuning, with 4 GPUs

CUDA_VISIBLE_DEVICES=0,1,2,3 python -u -m torch.distributed.run --nproc_per_node=4 \
--master_port=8889 main.py --distributed --config-base configs/RCAN/RCAN_Improved.yaml \
--config-file configs/RCAN/RCAN_x2_LP.yaml SOLVER.ITERATION_RESTART True SOLVER.ITERATION_TOTAL 40000 \
SOLVER.BASE_LR 8e-4 MODEL.PRE_TRAIN baseline/RCAN_x2_baseline/model/model_latest.pth.tar

Please change the checkpoint to your pretrained model path. Let me know if you have additional questions!

thanks