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

Numerial Stability Issue for MultiDecoder #15

Open
SwagJ opened this issue Aug 22, 2022 · 1 comment
Open

Numerial Stability Issue for MultiDecoder #15

SwagJ opened this issue Aug 22, 2022 · 1 comment

Comments

@SwagJ
Copy link

SwagJ commented Aug 22, 2022

Hi @stepankonev,

When I attempted to train the multi-decoder version with default config (final_RoP_Cov_A_fMCG.yaml), the traing process would be interrputed by infinity error at first epoch.
Setting trainable_cov to False will not help solve it.
If chaning running_mean_mode to "sliding", the training loss will be huge negative, and it becomes quite unstable

Did this happen when you train the network? And is this normal? Thank you in advance

@hushunda
Copy link

hushunda commented Oct 2, 2022

i have meet same problem. Have you deal with it? Can you share your resolution?
Thank you

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