Closed tangbohu closed 5 years ago
BTW, did you use SGD on these experiments reported in the paper.
I also have this question about what is the batch_size number??
HELLO: @tangbohu could we change e-mails to each others?? my email is paul8301526@gmail.com Maybe we could discuss this paper?? My wechat ID: paul8301526 HELLO: I found he use batchsize=128 when use the GPU(Titan X)
Hi, Indeed, it's 128: https://github.com/PatrykChrabaszcz/Imagenet32_Scripts/blob/master/WRNs_imagenet.py#L318
It's SGD with momentum: https://github.com/PatrykChrabaszcz/Imagenet32_Scripts/blob/master/WRNs_imagenet.py#L297
Hi, What batch size did you set on these experiments? Thanks