issues
search
fadel
/
pytorch_ema
Tiny PyTorch library for maintaining a moving average of a collection of parameters.
MIT License
400
stars
25
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
question about the use of ExponentialMovingAverage
#15
sevennotmouse
opened
3 weeks ago
2
Is the EMA used for training or just for tracking the EMA
#14
alisonpeard
opened
1 year ago
1
Running mean and running var use EMA?
#13
zhaohui-yang
opened
1 year ago
0
How to use it in pytorch-lightning?
#12
Devoe-97
opened
2 years ago
5
is model.eval() necessary within or outside context manager
#11
abaqusstudent
opened
2 years ago
1
Usage of weakref broke pickling/unpickling
#10
toshas
closed
2 years ago
4
Code does not work if some parameters are not trainable.
#9
GBATZOLIS
closed
2 years ago
3
Context manager & README updates
#8
Linux-cpp-lisp
closed
3 years ago
0
Device and dtype improvements
#7
Linux-cpp-lisp
closed
3 years ago
0
State dict support
#6
Linux-cpp-lisp
closed
3 years ago
1
Option to keep parameters reference in `ExponentialMovingAverage`
#5
Linux-cpp-lisp
closed
3 years ago
7
Tests and miscellaneous
#4
Linux-cpp-lisp
closed
3 years ago
1
Whether the model with ema needs bo be fixed in the validation phase —— “model(with ema).eval()”?
#3
TIanCat
closed
3 years ago
2
Add Feature: restore
#2
Zehui-Lin
closed
3 years ago
2
issues about add with torch.no_grad():
#1
liang233
closed
4 years ago
1