* extend description of train_model function

* fix m2mlstm & ndt data pre-formatting

neuro_py/ensemble/decoding/pipeline.py

@@ -234,6 +234,7 @@ def preprocess_data(hyperparams, ohe, nsv_train, nsv_val, nsv_test, bv_train, bv
     bins_before = hyperparams['bins_before']
     bins_current = hyperparams['bins_current']
     bins_after = hyperparams['bins_after']
+    is_2D = nsv_train[0].ndim == 1
     if hyperparams['model'] not in ('M2MLSTM', 'NDT'):
         (
             X_cov_train, X_flat_train, y_train,
@@ -262,6 +263,10 @@ def preprocess_data(hyperparams, ohe, nsv_train, nsv_val, nsv_test, bv_train, bv
             num_workers=hyperparams['num_workers'], modeltype=hyperparams['model'])
         hyperparams['model_args']['in_dim'] = X_train.shape[-1]
     else:
+        if is_2D:
+            nsv_train, bv_train = [nsv_train], [bv_train]
+            nsv_val, bv_val = [nsv_val], [bv_val]
+            nsv_test, bv_test = [nsv_test], [bv_test]
         if type(bv_train[0]) is pd.DataFrame:
             y_train = [y.values[:, hyperparams['behaviors']] for y in bv_train]
         else:
@@ -391,6 +396,9 @@ def shuffle_nsv_intrialsegs(nsv_trialsegs):
 def train_model(partitions, hyperparams, resultspath=None, stop_partition=None):
     """Generic function to train a DNN model on the given data partitions.
 
+    In-built caching & checkpointing is used to save the best model based on the
+    validation loss.
+
     Parameters
     ----------
     partitions : array-like