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

example MNIST does not work #2

Open
LeoFanCE opened this issue Jan 19, 2022 · 2 comments · May be fixed by #3
Open

example MNIST does not work #2

LeoFanCE opened this issue Jan 19, 2022 · 2 comments · May be fixed by #3

Comments

@LeoFanCE
Copy link

Hello,
Thank you for your code. While I was running your MNIST example an error appeared.
error

I added statements in your code to prevent the error since the truth value of an array with more than one element is ambiguous.
add

But as a result, the model could not converge like that you showed in your github page.
result1

I wonder if you had the same problem or is there something that I've missed.

@gcamfer gcamfer linked a pull request Jan 20, 2022 that will close this issue
@gcamfer
Copy link
Owner

gcamfer commented Jan 20, 2022

Hi, thank you very much for trying it and finding issues.
I don't really know why I returned a list of booleans, I guess my head was thinking about other stuff.
I have fixed that problem.

Regarding the convergence, it seems to be a hyperparameter problem. Try reducing the exploration or reducing the learning rate. You can even rerun as random initialisation will get you to another point, but it's not very stable.

You can try the branch in this Colab and next I'll merge with master.

Greetings

@JuiMahind
Copy link

Hello,
After execution of the code I am getting this error

UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use the zero_division parameter to control this behaviour.

this error is resulting in no f1 score for some labels.
can you give me any solution for this error

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

Successfully merging a pull request may close this issue.

3 participants