-
Notifications
You must be signed in to change notification settings - Fork 194
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
Loss NaN's #5
Comments
Fixed this by replacing 1e-8 in the log()'s with 1e-6... but the AdamOptimizer's behavior is quite weird, the training goes well for some time and then the loss blows up for no reason. Looking at the results before the loss blows up, it clearly gets worse. Maybe this is just part of stochastic GD.
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Anyone else have this problem?
The images decoded then go "dark" right after that.
The text was updated successfully, but these errors were encountered: