Stacy Yang

just ping the dataset organizer

checked with the author @zhanghang1989 , the next version should be released in 2~3 weeks, which will support single gpu training.

> @zhanghang1989 GC = tmp.sum(0).sum(0) in https://github.com/zhanghang1989/PyTorch-Encoding/blob/master/encoding/lib/cpu/encoding_cpu.cpp#L44 should be GC = -tmp.sum(0).sum(0) ? The weights can learn the negative value automatically if necessary.

Please read the docs on PyTorch for customizing models.

FYI PyTorch has supported SyncBN officially. The author is not maintaining this repo and has moved to MXNet