Commit History

Created using Colab
f34a470

ribesstefano commited on

Added tutorial to be run in Colab
db62a2a

ribesstefano commited on

Fixed minor bug when returning predictions
dcdef4a

ribesstefano commited on

Added support for listing available dataset data
42d3d55

ribesstefano commited on

Fixed bug on averaging predictions in package code
367cf2c

ribesstefano commited on

Fixed predictions moving to CPU from GPU in package code
8d92d67

ribesstefano commited on

Fixed rdkit requirements in setup.py
8489d4a

ribesstefano commited on

Added version requirement on rdkit setup.py
dac3be7

ribesstefano commited on

Fixed code on loading cell embeddings for package
c06df22

ribesstefano commited on

Fixed bug on loading cell embeddings in package code
e21cad4

ribesstefano commited on

Fixed loading cell embeddings in package code
229284f

ribesstefano commited on

Fixed bug when loading cell2embeddings from wrong directory
9bd496b

ribesstefano commited on

Added XGBoost majority vote performances
a8d1800

ribesstefano commited on

Added notebook for plotting results
d362c4b

ribesstefano commited on

Added notebook on cell embeddings generation + Added XGBoost reports
8315295

ribesstefano commited on

Added XGBoost Optuna training + Added Ablation studies with zeroed input vectors
33f1644

ribesstefano commited on

Changed ablation to input zero-valued vactors
4bf0ec2

ribesstefano commited on

Added results with FP normalization removed
d8a186f

ribesstefano commited on

Added Maturity level badge
9e3ebaa

ribesstefano commited on

Added ablation study and performance plots
c8d51e3

ribesstefano commited on

Added ablation study with randomly sampled vectors + Started working on LightningDataModule wrapper
4e1d3f6

ribesstefano commited on

Added convincing results (really, before it was a fast_dev_run)
48aea13

ribesstefano commited on

Added convincing results
b7582e0

ribesstefano commited on

Added softmax in model + Fixed and updated some hparams + Fixed bug in tanimoto distance (it was treated as similarity)
bda3015

ribesstefano commited on

Added FP search + Added train metrics to logs
fda7af7

ribesstefano commited on

Added LR scheduler + set default sum of embeddings
62ccb16

ribesstefano commited on

Added majority voting evaluation
b86d3ec

ribesstefano commited on

Improved plotting code
74a86c6

ribesstefano commited on

Pushing current results
9692a92

ribesstefano commited on

Working loading and predicting with pretrained model + Added scaling in "pure" pytorch during forwarding
15216c3

ribesstefano commited on

Improved logging + current optuna studies
0171744

ribesstefano commited on

Added model tests + checkpointing of the scaler object
de956c8

ribesstefano commited on

Refactored experiments + fixed bug in dataset when applying scaling to val and test sets
6a5a99e

ribesstefano commited on

Updated protac dataset to handle missing values in embedding dictionaries
4d17fea

ribesstefano commited on

refactored package code and fine running experimental code
165d38a

ribesstefano commited on

added rdkit to setup
74be897

ribesstefano commited on

started testing package code
ea572f9

ribesstefano commited on

Started working on packaging the repository
5e01175

ribesstefano commited on

Updated 10% test with fixed Tanimoto-based split
6101de8

ribesstefano commited on

Updated reports with ablation studies w/out E3
b496192

ribesstefano commited on

Started working on optuna-based sklearn models training
f9a730b

ribesstefano commited on

Added notebook on predicting degradation with sklearn models
f258949

ribesstefano commited on

Added IaP to IAP conversion in data curation
e7f9851

ribesstefano commited on

Added save and restore of Optuna studies
dafeec2

ribesstefano commited on