Comments (4)
I replace the dataset with Cifar10 ,and reduce the number of layers.When i training this model ,the accuracy rate maintain 10%.It can't change
As far as I am concerned, I think may be the learning rate is to blame which is different from the official version and may be not appropriate. Additionally, the initial state is very important. Maybe you can compare this code with the official version and replace that dataset with CIFAR-10 to have a try.
from single-path-one-shot-nas.
I replace the dataset with Cifar10 ,and reduce the number of layers.When i training this model ,the accuracy rate maintain 10%.It can't change
As far as I am concerned, I think may be the learning rate is to blame which is different from the official version and may be not appropriate. Additionally, the initial state is very important. Maybe you can compare this code with the official version and replace that dataset with CIFAR-10 to have a try.
Sorry, have you tried training your model on the imagenet dataset? I ran the CIAFAR10 dataset with the official code, and the loss will be reduced! I am confused and need your help.
from single-path-one-shot-nas.
I replace the dataset with Cifar10 ,and reduce the number of layers.When i training this model ,the accuracy rate maintain 10%.It can't change
As far as I am concerned, I think may be the learning rate is to blame which is different from the official version and may be not appropriate. Additionally, the initial state is very important. Maybe you can compare this code with the official version and replace that dataset with CIFAR-10 to have a try.Sorry, have you tried training your model on the imagenet dataset? I ran the CIAFAR10 dataset with the official code, and the loss will be reduced! I am confused and need your help.
Maybe you can contact me via [email protected] for detail and we can have a further talk.
from single-path-one-shot-nas.
This issue is caused by the lack of relu and it has been updated.
from single-path-one-shot-nas.
Related Issues (16)
- How to do channel search? HOT 1
- 请问下,在block selection时,不同的branch之间会共享权重么? HOT 1
- model test HOT 1
- net output dimension mismatch on cifar10 HOT 7
- about extracting final result HOT 2
- 您好,关于精度复现 HOT 3
- Can not find cifar_train.py HOT 2
- Why shuffle_channels function is not used? HOT 2
- 请问最终的最高精度能达到多少? HOT 1
- 采样出一条路径后,需要对其他的路径冻结梯度更新吗? HOT 2
- 根据经验,cifar10和imagenet对应的哪些超参需要调整? HOT 1
- Randomly sample and save models HOT 4
- 运行完 supernet.py 后,如何进行模型加载校验(val)? HOT 4
- BN校准 HOT 2
- 关于Mixed-Precision Quantization HOT 1
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 single-path-one-shot-nas.