You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello @ilyes319,
I was wanting to use intermediate layers from MACE-MPA-0 (frozen) and decompose them into individual irreps for creating custom heads on top for full dielectric tensor predictions, I wanted to use the latest version of e3nn which gives an option to filter the irreps directly. I see all the mace-torch have used e3nn 0.4.4. Any plans on upgrading that to the latest?
Aside, do you have any suggestions on which pretrained mace could be best for this task even though its not directly trained on polarization data?
Thanks!
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
-
Hello @ilyes319,
I was wanting to use intermediate layers from MACE-MPA-0 (frozen) and decompose them into individual irreps for creating custom heads on top for full dielectric tensor predictions, I wanted to use the latest version of e3nn which gives an option to filter the irreps directly. I see all the mace-torch have used e3nn 0.4.4. Any plans on upgrading that to the latest?
Aside, do you have any suggestions on which pretrained mace could be best for this task even though its not directly trained on polarization data?
Thanks!
Beta Was this translation helpful? Give feedback.
All reactions