Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / joe-siyuan-qiao/weightstandardization issues and pull requests
#29 - style(readme): shorten pytorch implementation
Pull Request -
State: open - Opened by ClashLuke over 2 years ago
#28 - WS with ConvTranspose2d?
Issue -
State: open - Opened by hiyyg about 3 years ago
#27 - How to apply WS in video understanding?
Issue -
State: closed - Opened by checheche222 over 3 years ago
#26 - Implementation question in Tensorflow
Issue -
State: open - Opened by jbp70 almost 4 years ago
- 4 comments
#25 - Possible typo in the paper and some other questions
Issue -
State: open - Opened by zhujiagang about 4 years ago
#24 - consistent mean calculation with std calculation
Issue -
State: open - Opened by huangeddie about 4 years ago
- 2 comments
#23 - It just explodes!!!
Issue -
State: open - Opened by VCasecnikovs over 4 years ago
- 3 comments
#22 - How to load weights of Weight Standardization model?
Issue -
State: open - Opened by mobassir94 over 4 years ago
#21 - Error with multi_gpu_model on Keras
Issue -
State: open - Opened by Xonxt over 4 years ago
#20 - Network's performance decreases after adopting WS
Issue -
State: open - Opened by NoOneUST over 4 years ago
- 5 comments
#19 - WD or WD+GN with fp16
Issue -
State: open - Opened by vadik6666 over 4 years ago
- 3 comments
#18 - It seems that same method has been proposed before
Issue -
State: open - Opened by netw0rkf10w almost 5 years ago
- 2 comments
#17 - AttributeError: 'module' object has no attribute 'reduce_mean'
Issue -
State: closed - Opened by longmalongma almost 5 years ago
- 4 comments
#16 - changing the position of epsilon
Issue -
State: open - Opened by implus about 5 years ago
- 2 comments
#15 - tf.math.reduce_mean(kernel, axis=[0, 1, 2],
Issue -
State: closed - Opened by henbucuoshanghai about 5 years ago
- 2 comments
#14 - Can i apply the same method to 3D conv?
Issue -
State: open - Opened by seominseok0429 about 5 years ago
- 5 comments
#13 - Using BN+WS in ResNet50 similar or slightly worse than using BN only in ResNet50 for ImageNet classification
Issue -
State: closed - Opened by niam06 about 5 years ago
- 4 comments
#12 - Possible Keras Implementation doesn't seem to work
Issue -
State: closed - Opened by pgadosey about 5 years ago
- 7 comments
#11 - It seems to have no effect on this work.
Issue -
State: open - Opened by MoonWolf9067 about 5 years ago
- 1 comment
#10 - WS of Deformable Convolution.
Issue -
State: open - Opened by lizhen2017 over 5 years ago
- 3 comments
#9 - batch size and iteration
Issue -
State: open - Opened by pursueorigin over 5 years ago
- 2 comments
#8 - the loss is nan
Issue -
State: open - Opened by Maycbj over 5 years ago
- 6 comments
#7 - About pretrained model R101-GN-WS
Issue -
State: closed - Opened by JrPeng over 5 years ago
- 4 comments
#6 - In my model, GN+WS is worse
Issue -
State: closed - Opened by DrewdropLife over 5 years ago
- 4 comments
#5 - about train and inference
Issue -
State: open - Opened by lichunxiang93 over 5 years ago
- 1 comment
#4 - Quesions on accuracy gain of WS?
Issue -
State: closed - Opened by huangh12 over 5 years ago
- 2 comments
#3 - In my test,gn+ws result is worse
Issue -
State: closed - Opened by hust-kevin over 5 years ago
- 2 comments
#2 - Conv1d Version?
Issue -
State: closed - Opened by kayuksel over 5 years ago
- 5 comments
#1 - Questions about the definition of WS and a possible typo in the paper.
Issue -
State: closed - Opened by t-taniai over 5 years ago
- 7 comments