Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

I want to test the quantitative effect, what do I need to do? #5

Open
suhong1 opened this issue Oct 18, 2019 · 1 comment
Open

I want to test the quantitative effect, what do I need to do? #5

suhong1 opened this issue Oct 18, 2019 · 1 comment

Comments

@suhong1
Copy link

suhong1 commented Oct 18, 2019

No description provided.

@TF2-Engine
Copy link
Owner

Assuming you have downloaded the compressed weight: https://1drv.ms/u/s!Am9Mk04MA_K1akWOEe6IRlw_ons?e=02ApjG and the corresponding dataset and readed the README of the quantization part of the Transform_kit. Then you should have created the activations Q file, and you could test the inference accuracy on ImageNet or other dataset. Currently, you can modify the code under https://github.com/TF2-Engine/TF2/tree/master/Runtime_Engine/cnn/host to the CNN that you want test, and combine the Q file of different layers those you need and then 'make', run the 'run.sh' file. You can read the details of the host code and it will help a lot if you want to test more.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants