lumiaomiao

> Hi, sorry about that. Wasabi cloud has a rule of limited downloading volume. I am looking for other solutions @zhanghang1989 I ran into the same problem. You could use...

> Hi @DAVID-Hown could you try https://dl.dropboxusercontent.com/s/fi72lae1exon3o4/resnest50-528c19ca.pth?dl=0 @zhanghang1989 Could you update the link of https://s3.us-west-1.wasabisys.com/resnest/detectron/faster_cascade_rcnn_ResNeSt_101_FPN_syncbn_range-scale_1x-3627ef78.pth please?

> Maybe you can try using pytorch==1.4.0 or pytorch==1.2.0 nan doesn‘t disappear in pytorch==1.4.0, and I can't configure pytorch==1.2.0.

Thank you for your replay.

@zhihou7 Hi, I have another question about the code. The function get_new_Trainval_N in lib/ult/ult.py is definied as :  Why use " Trainval_N[4]" not " Trainval_N[k]" ?

Thank you for your quick reply.

@zhihou7 As following codes, if an image contains two pairs , , and the first one is in the unseen composition list, then you delete two pair from training data....

I get it, thank you.

Hi, I sent an applications two weeks ago but no answer so far. Could you send me a download link? My email is [email protected]. Thank you

> @lumiaomiao > Hi, > > I sent the link yesterday. You can check your email if you received it. > > tangsipeng Hi, I have received the email, thank...