image2video-synthesis-using-cINNs icon indicating copy to clipboard operation
image2video-synthesis-using-cINNs copied to clipboard

Train own dataset, load cINN model.'state_dict' is error

Open bjz0926 opened this issue 2 years ago • 0 comments
trafficstars

mod stage2_cINN/main.py file,
if PFVD < best_PFVD: torch.save(save_dict, save_path + '/checkpoint_best_val.pth')


if PFVD < best_PFVD: torch.save(network.state_dict(), save_path + '/checkpoint_best_val.pth')

by python -W ignore -m stage2_cINN.main -gpu <gpu_id> -cf stage2_cINN/configs/ I get the checkpoint_best_val.pth. I change its name to cINN.pth, but error. Please tell the main.py that is right or error. Thank you

bjz0926 avatar Sep 12 '23 02:09 bjz0926