[ENH] Neat and automated transfer learning with OPTIMADE API for auto-adjusted problem-specific ML model generation on the fly#16

Merged

amkrajewski merged 50 commits intomainfrom add_ModelTuner

Apr 4, 2024

+1,163-2

Commits on Mar 28, 2024

(MA) improved memory handling during adjustments
amkrajewski
committed
(MA) implemented a hyperparameter search matrix for 27 common options; meant mostly for tuning to smaller datasets
amkrajewski
committed
(MA) implemented a hyperparameter search matrix for 27 common options; meant mostly for tuning to smaller datasets
amkrajewski
committed
(MA) added optimade to the dev requirements
amkrajewski
committed
(MA) implemented initialization of OPTIMADEAdjuster and its docstrings
amkrajewski
committed
(MA) implemented querying OPTIMADE and passing structures to featurization inOPTIMADEAdjuster
amkrajewski
committed
(core) added [http_client] to the optimade requirement needed for querying
amkrajewski
committed
(MA) added entry name collection to the OPTIMADEAdjuster
amkrajewski
committed
(MA) upped the default result number limit to 10000 and added it as a parameter with documentation in OPTIMADEAdjuster
amkrajewski
committed
(core) removed info about importing from the top pysipfenn namespace in the top __init__
amkrajewski
committed
(MA) finalized the fetchAndFeturize routine of OPTIMADEAdjuster, now collecting names, target data, and featurizing the obtained structures
amkrajewski
committed
(MA) add a message on possible degeneracy of datasets from overlapping queries
amkrajewski
committed
(MA) added class parameter of validationLabels, assigned at every adjustment, to display if a datapoint has been shown to the model at adjustment. Displayed when plotting unadjusted and adjusted…
amkrajewski
committed
(MA) improved several printouts
amkrajewski
committed
(MA) added assertions that data to operate on is available before tuning
amkrajewski
committed

Commits on Mar 29, 2024

Commits on Mar 30, 2024