Gavin666Github

Have anyone come accross this phenomenon? size of my training datasets is 20000 pairs my configuration with params: --batchSize 64 --no_html --nThreads 16 --niter 800 --niter_decay 8000 --epoch_count 729 --gpu_ids...

@zhujingsong hi,jing song,"We use list files in data/ucf101/ subdir to make the code find RGB images and flow data saved on disk" ,have you solved the problem? if YES,please share...

@XudongLinthu hello,xu,I'am in China but i can't download the link shared.would you please share the datasets  with baidu cloud disk to me?many thanks！

@Rhythmblue aha~~ It's so big. It's going to be very slow to download. Thank you for your advice.I could try.

我也遇到了这个错误，请问解决了吗

我没有训练baseline，直接开始稀疏化训练，有问题吗

@JimXiongGM 你说的太对了，超分这个课题都是这么玩的，从高清图按一定方法下采样得到LR图，然后再训练下返回得到高清图，再算个psnr 、ssim什么的刷分数，OK可以提交paper了，问题是这些对数据有严重的依赖性，对下采样方法也有局限和依赖，事实上真正低分辨率的图片本身信息量不足，很可能不是按照常规下采样方法就能得到同样质量的照片，问题就在这里，用这个LR超分的时候，得到的HR可能就会气死你

netG = net.dehaze(inputChannelSize, outputChannelSize, ngf) 报错从这开始，上述代码放在这里没用的

@SumiHui @hezhangsprinter Thank you very much! I believe he will be happy and willing for you to share the link to others because it's benefit to more people.

mutual learling 需要在两个GPU跑，一个跑不起来,内存没有及时释放，慢慢的显存爆了