Fix construction of ModelListGP in closed_loop_botorch tutorial (#175)

Summary:
Resolves small issue flagged in #174
Pull Request resolved: #175

Reviewed By: sdaulton

Differential Revision: D15861130

Pulled By: Balandat

fbshipit-source-id: e137d14041a5eaad31f4eec729f872badfd4325c

Author: Balandat

tutorials/closed_loop_botorch_only.ipynb

@@ -102,7 +102,7 @@
     "    model_obj = FixedNoiseGP(train_x, train_obj, train_yvar.expand_as(train_obj)).to(train_x)\n",
     "    model_con = FixedNoiseGP(train_x, train_con, train_yvar.expand_as(train_con)).to(train_x)\n",
     "    # combine into a multi-output GP model\n",
-    "    model = ModelListGP(gp_models=[model_obj, model_con])\n",
+    "    model = ModelListGP(model_obj, model_con)\n",
     "    mll = SumMarginalLogLikelihood(model.likelihood, model)\n",
     "    # load state dict if it is passed\n",
     "    if state_dict is not None:\n",