Skip to content

Releases

v1.0.1 — model weights

Choose a tag to compare

@minotm minotm released this 21 Jun 09:59

Pretrained model weights for the active-drug-discovery reference implementation.

Download into src/model_weights/ (see the README "Downloading weights and data" section):

  • equiformer_regio.pth — regioselectivity GNN (EquiformerV2)
  • ponita_binary.pth — binary reactivity GNN (PONITA)
  • binary_xgb.zip — binary reactivity XGBoost ensemble + preprocessing; unzip into src/model_weights/binary_xgb/

No Minisci weights are included; train your own (see the README "Minisci training" section).

v1.0.0 — model weights

Choose a tag to compare

@minotm minotm released this 21 Jun 09:45

Pretrained model weights for the active-drug-discovery reference implementation.

Download into src/model_weights/ (see the README "Downloading weights and data" section):

  • equiformer_regio.pth — regioselectivity GNN (EquiformerV2)
  • ponita_binary.pth — binary reactivity GNN (PONITA)
  • binary_xgb.zip — binary reactivity XGBoost ensemble + preprocessing; unzip into src/model_weights/binary_xgb/

No Minisci weights are included; train your own (see the README "Minisci training" section).