added deprecation note

some updates to comply with newer API

move the CV args to the BrainScore init call

Update test_mean_froi_pereira2018_firstsessions.py

Merge branch 'feature-parallelize-mapping-1' of github.com:language-brainscore/lang-brainscore-fuzzy-potato into feature-parallelize-mapping-1

brainscore methods no longer take split coords as args; it is part of the object

brainscore methods no longer take split coords as args; it is part of the object

gitignore wandb logs

allow tokenizers parallellelism explicitly

make sample_ and neuroid_split_coord part of Mapping and BrainScore objects; they were previously not tracked

use joblib in mapping fit_transform for learning a model per neuroid

handle params that are dicts; e.g. hparams of a ridgecv

bugfix: need `.cpu()` before `.numpy()` when running on GPU

use wandb logging utils functions to log params to wandb

write helper wandb logging methods in logging utils