google-deepmind / google-deepmind/weathernext
Removing receiver conditioning in graph network
Open
- Dominant language
- Python
- Stars
- 7.7k
- Forks
- 986
- PR merge metrics
- No merged PRs in 30d
Description
I'm trying to remove conditioning on receiver mesh modes in the edge model of the grid2mesh encoder (as mentioned in the FGN paper). However, I'm noticing quite a few of artifacts in longer rollouts.
In `typed_graph_net.py`, I removed the conditioning in both the `_edge_update` function at the `InteractionNetwork` instantiation. Is there anything else I'm missing? Any code examples would be extremely helpful!
Contributor guide
Assessment
This issue has not been assessed yet.