xwd-szu

the same problems, Did you find a solution?

run with only one Gpu , a problem of diffrent input image size run with two Gpus , one GPUS input image size = (1,3,335,500) while another GPUS input image...

ok！thank to your answer！however，how i train with my own datasets，and what mean under dataloaders TXT file ，is there any py_file to generate these txt_file?

I re-pulled your code, all settings have not changed. I am running on two A100 GPUs, but the result is still not ideal.

Thanks for your reply！ I checked the problem you mentioned and ran it a few more times, now it meets the metrics of the paper on the Sintel dataset, but...

i means feat_q and feat_k dim ，is B * C * W * H？or other？and each dim means？

thanks for your answer.  this feat_id is means?  the output dim of this netG？and each dim means？ Look forward to your answer！！

Can you provides the code of T-SNE of features visualization? thank you very much!!!

Have you found this good implementation？

> problem i still can't sovel it