-
Notifications
You must be signed in to change notification settings - Fork 16
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
[Feature] Add training strategies based on learning Tendencies #47
Labels
enhancement
New feature or request
Comments
Rilwan-Adewoyin
changed the title
[Feature] Tendency based optimization
[Feature] Add training strategies based on Tendency learning
Sep 4, 2024
Rilwan-Adewoyin
changed the title
[Feature] Add training strategies based on Tendency learning
[Feature] Add training strategies based on learning Tendencies
Sep 4, 2024
Rilwan-Adewoyin
added a commit
that referenced
this issue
Sep 4, 2024
Rilwan-Adewoyin
added a commit
that referenced
this issue
Sep 4, 2024
Rilwan-Adewoyin
added a commit
that referenced
this issue
Sep 4, 2024
…earning state Co-authored-by: Jakob Schloer <jakob.schloer@gmail.com>
Rilwan-Adewoyin
added a commit
that referenced
this issue
Sep 4, 2024
Co-authored-by: Jakob Schloer <jakob.schloer@gmail.com>
Rilwan-Adewoyin
added a commit
that referenced
this issue
Sep 4, 2024
Co-authored-by: Jakob Schloer <jakob.schloer@gmail.com>
Rilwan-Adewoyin
added a commit
that referenced
this issue
Sep 4, 2024
Co-authored-by: Jakob Schloer <jakob.schloer@gmail.com>
Rilwan-Adewoyin
added a commit
that referenced
this issue
Sep 4, 2024
Co-authored-by: Jakob Schloer <jakob.schloer@gmail.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Is your feature request related to a problem? Please describe.
@Rilwan-Adewoyin > Allow optional training with optimisation towards tendency as opposed to state/residual
@jakob-schloer > While the deterministic AIFS is currently trained on the weighted mean square error of absolute values, GraphCast is trained to predict the normalized tendency. This might be beneficial for variables changing on different timescales.
Describe the solution you'd like
Instead of predicting x(t+1) directly, the model should predict the normalized tendency y(t) = x(t+1) - x(t).
Describe alternatives you've considered
NA
Additional context
Has been shown to improve results with a diffusion model targeting atmosphere - this will be useful for investigating whether or not improvements extend to other models and other domains such as ocean and wave.
Dependent on a issue/branch on anemoi-models repository: ecmwf/anemoi-models#22
@jakob-schloer
@da-ewanp
@ssmmnn11
Organisation
ECMWF
The text was updated successfully, but these errors were encountered: