issues
search
WangYZ1608
/
Knowledge-Distillation-via-ND
The official implementation for paper: Improving Knowledge Distillation via Regularizing Feature Norm and Direction
13
stars
2
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
There are something wrong in nd_loss compute
#10
qiuxiaqing
opened
1 year ago
2
Is the teacher model provided by CIFAR corrupted?
#9
liuyyy111
opened
1 year ago
1
weight of 3 losses for bigger teachers
#8
HanGuangXin
closed
1 year ago
2
Could you kindly provide the training log?
#7
HanGuangXin
closed
1 year ago
1
why teacher model and student model have different operation?
#6
HanGuangXin
closed
1 year ago
1
use only 1 gpu per node?
#5
HanGuangXin
closed
1 year ago
2
ValueError: too many values to unpack (expected 2)
#4
HanGuangXin
closed
1 year ago
2
train_kd.py: error: unrecognized arguments: --norm_loss_factor 1.0
#3
HanGuangXin
closed
1 year ago
3
How to download teacher and student weights?
#2
HanGuangXin
closed
1 year ago
3
How to reproduce the center_emb_train.json file for ViT model?
#1
onvungocminh
closed
1 year ago
1