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

problem with training for loss #18

Closed
iWeisskohl opened this issue Apr 26, 2018 · 6 comments
Closed

problem with training for loss #18

iWeisskohl opened this issue Apr 26, 2018 · 6 comments

Comments

@iWeisskohl
Copy link

HI@ @suhangpro ,
Thanks for your code. But when I run this code,I only got loss=87.33. I have checked the label and input ,they are right. I am wondering is there some problem with the batch size in data layer when your definite it ? Because I try to set batch_size as 32, but it meets some problem.

@Hwang64
Copy link

Hwang64 commented May 8, 2018

@iWeisskohl I meet the same question, have you solved this problem?

@iWeisskohl
Copy link
Author

@Hwang64 Not yet..... there might be some problem with batch size , but I can't figure it out with the offered code.

@Hwang64
Copy link

Hwang64 commented May 18, 2018

@iWeisskohl , refer to the tensorflow version, the learning rate should be set to 0.0001 but not 0.01 and I get the accuracy which is almost the same as the paper illustrated, you may have a try.

@iWeisskohl
Copy link
Author

@Hwang64 Thanks so much.I got the almost same results. But I have one more question, have u ever pay attention about the training data and test data? It seems the offered data is not as same size as it reported in the paper....it only contains about 3000 training and 800 testing,but it supposed to have the same size with http://3dshapenets.cs.princeton.edu/ ModelNet40.zip

@TimeIvyace
Copy link

actually i meet the same problem and i find that through modify the batch_size this issue can be solved well

@suhangpro
Copy link
Owner

Thanks all for contributing to the solution!

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

4 participants