On Sam's gail
branch, I checked whether the following scripts would fail as expected when replacing arbitrary Sacred configuration options with a bad value.
pretrain_n_adapt.py
(calledchain_ex_bc.pkl
in results files)run_rep_learner.py
(calledrepl_{algo}.pkl
in results files)il_train.py
(calledil_train_test_ex_{algo}.pkl
in the results files)il_test.py
(calledil_train_test_ex_{algo}.pkl
in the results files)
The bad value that I used in particular was ...
, which satisfies bool(...) == True
and ... is not None
.