2021-NeurIPS-NCR

2021-NeurIPS-NCR copied to clipboard

question on the loss

1

https://github.com/XLearning-SCU/2021-NeurIPS-NCR/blob/c0ea996cba4e23cc1f6b069e638463d6c4cb495f/NCR/model.py#L484 Line 484 and Line 486 in model.py. Should it be "margin" instead of "self.margin" since you have applied soft margin.

haochenglouis

疑似存在代码实现错误

1

co-train.py中： ptl = batch_prob_l * batch_labels_l + (1 - batch_prob_l) * pl[0] 中pl[0]仅是batch内第一个伪标签，这里应该是和batch内所有的进行累加。 希望能得到您的回复。

sillytheif