Skip to content

Conversation

sgreenbury
Copy link
Collaborator

@sgreenbury sgreenbury commented Sep 25, 2025

Closes #855.
Closes #856.

(builds on changes in #840)

This is WIP exploration of adding further API to TheWellEmulator base class that would support more complex training and losses.

  • Add learnable weights for prediction ahead in free running rollout
  • Add teacher forcing scheduling
  • Add support for FNO with temporal FT
  • Add example of CNN encoder
  • Add prototype training script and config to run from CLI

Copy link

Check out this pull request on  ReviewNB

See visual diffs & provide feedback on Jupyter Notebooks.


Powered by ReviewNB

- Check perf compared with The Well (VMRSE and MSE driving diffs)
- Add The Well emulator without rehsaping time
- Refactor with optional time and constants in the API
- Update loss fn to be passed from the The Well metrics for comparison
- Update notebook
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 this pull request may close these issues.

1 participant