YirongMao/softmax_variants

Issues

• Why is LGMLoss `margin_logits` is equivalent eq.17

  #9 opened 2 years ago by Chen-Dixi
  0

• LGM loss训练时出现loss为负值

  #8 opened 3 years ago by liuge1008
  0

• loss.data[0] gives an error instead use loss.item()

  #5 opened 4 years ago by kalyanainala
  0

• ValueError: optimizer got an empty empty parameter list

  #7 opened 4 years ago by liuge1008
  0

• what is feat_dim in LGMLOSS_v0,How can I set it

  #2 opened 6 years ago by AnnaRose0226
  2

• In class LGMLoss(nn.Module), which is the implementation of the original paper the covariance was considered as identity matrix instead of a diagonal matrix

  #6 opened 4 years ago by kalyanainala
  0

• How to understand the centers shown in the LMCL_loss?

  #4 opened 5 years ago by taylover-pei
  1

• can you implement the affinity-loss

  #3 opened 5 years ago by ray-lee-94
  2

• Margin in LMCL loss might have some problems.

  #1 opened 6 years ago by ZhangXLaurence
  1