Comments (4)
yzd-v
commented on August 26, 2024
1
就应该是1-Tt，因为softmax的缘故，non-target之和始终为1-Tt，我把Tt人为的缩小（可能很接近0了）是为了让1-Tt不等于0，从而避免nan
from cls_kd.
yzd-v
commented on August 26, 2024
It should be 0 when calculating distributed loss. For soft loss, what we need is wt in line 32.
from cls_kd.
cotyyang
commented on August 26, 2024
您好，您的回复确实和我理解的是一致的，但是可能是我表达有错误。我这里意思是论文中公式5的分母T_t应该是target class probability，这个公式对应代码45-46行，但是代码42-43行中的T_t始终是0。所以45行的1-T_t应该是1-w_t吗
from cls_kd.
cotyyang
commented on August 26, 2024
谢谢解答，确实是我把这里代码理解错了，T_i总和本身就已经是1-T_t
from cls_kd.
Related Issues (20)
- Dataset question
- Cifar100 dataset question HOT 6
- Question about ViTKD loss code HOT 5
- Some question about ViTKD HOT 1
- About the ratio between Lori and Lsoft HOT 1
- Questions on masked area
- Can you provide the detection code？ HOT 1
- Positional Embedding HOT 2
- Welcome update to OpenMMLab 2.0
- Distillation 文件夹在哪呢? HOT 1
- loss is always nan HOT 1
- can you told me which code can generate figure 1 in paper? HOT 1
- Experimental results reproducibility issues HOT 1
- Implementation of non-target mask HOT 3
- Training DeiT + NKD with ViTKD HOT 3
- I don't understand where "Normalized" is reflected or special in the NKD code? HOT 2
- 在CIfar100上复现resnet18, USKD方法的问题 HOT 4
- About DeiT-Tiny model in your experiment for ViTKD
- 训练
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from cls_kd.