Remove HeterogeneousSingleTaskGP

Summary:
X-link: pytorch/botorch#2616

This model has been buggy for quite a long time and we still haven't fixed it. Removing it should be preferable to keeping around a known buggy model. Example bug reports:
- pytorch/botorch#861
- pytorch/botorch#933
- pytorch/botorch#2551

Differential Revision: D65543676

ax/models/tests/test_torch_model_utils.py

@@ -17,7 +17,7 @@
     tensor_callable_to_array_callable,
 )
 from ax.utils.common.testutils import TestCase
-from botorch.models import HeteroskedasticSingleTaskGP, SingleTaskGP
+from botorch.models import SingleTaskGP
 from botorch.models.deterministic import GenericDeterministicModel
 from botorch.models.model import ModelList
 from botorch.models.model_list_gp_regression import ModelListGP
@@ -31,9 +31,9 @@ def test_is_noiseless(self) -> None:
         x = torch.zeros(1, 1)
         y = torch.zeros(1, 1)
         se = torch.zeros(1, 1)
-        model = SingleTaskGP(x, y)
+        model = SingleTaskGP(train_X=x, train_Y=y)
         self.assertTrue(is_noiseless(model))
-        model = HeteroskedasticSingleTaskGP(x, y, se)
+        model = SingleTaskGP(train_X=x, train_Y=y, train_Yvar=se)
         self.assertFalse(is_noiseless(model))
         with self.assertRaises(ModelError):
             is_noiseless(ModelListGP())
@@ -181,7 +181,7 @@ def test_with_obj_thresholds_can_subset(self) -> None:
 
     def test_unsupported(self) -> None:
         yvar = torch.ones(1, 2)
-        model = HeteroskedasticSingleTaskGP(self.x, self.y, yvar)
+        model = SingleTaskGP(train_X=self.x, train_Y=self.y, train_Yvar=yvar)
         subset_model_results = subset_model(model, self.obj_weights)
         model_sub = subset_model_results.model
         obj_weights_sub = subset_model_results.objective_weights