Closed daixiangzi closed 4 years ago
It is two different augmentations on the same unlabeled data.
oh,you are right,thank you very much
---Original--- From: "YUI"<notifications@github.com> Date: Thu, Dec 19, 2019 19:00 PM To: "YU1ut/MixMatch-pytorch"<MixMatch-pytorch@noreply.github.com>; Cc: "daixiangzi"<543826458@qq.com>;"Author"<author@noreply.github.com>; Subject: Re: [YU1ut/MixMatch-pytorch] why unlabel numbers two times than label? (#26)
It is two different augmentations on the same unlabeled data.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.
(inputs_u, inputsu2), = unlabeled_train_iter.next() not should label numbers == unlabeled numbers?