fix task type for multilabel and multiclass classification when passe…

…d to UX (#15)

responsibleai_text/rai_text_insights/rai_text_insights.py

@@ -15,7 +15,6 @@
 from erroranalysis._internal.cohort_filter import FilterDataWithCohortFilters
 from ml_wrappers import wrap_model
 from raiutils.data_processing import convert_to_list, serialize_json_safe
-from raiutils.models import ModelTask as RAIModelTask
 from raiutils.models import SKLearn, is_classifier
 from responsibleai._interfaces import Dataset, RAIInsightsData
 from responsibleai._internal.constants import (ManagerNames, Metadata,
@@ -432,15 +431,13 @@ def _is_classification_task(self):
     def _get_dataset(self):
         dashboard_dataset = Dataset()
         tasktype = self.task_type
-        is_classification_task = self._is_classification_task
-        if is_classification_task:
-            tasktype = RAIModelTask.CLASSIFICATION
         if isinstance(tasktype, Enum):
             tasktype = tasktype.value
         dashboard_dataset.task_type = tasktype
         dashboard_dataset.categorical_features = []
         dashboard_dataset.class_names = convert_to_list(
             self._classes)
+        is_classification_task = self._is_classification_task
         dataset = self._get_test_without_target(is_classification_task)
 
         predicted_y = None