Execute notebooks end-to-end and fix typo in models.py

- Execute both notebooks (prediction_and_optimization_tutorial, strength_curve_prediction_demo) end-to-end in concrete mode so all cell outputs are populated, including Pareto frontier plots and qLogNEHVI acquisition function optimization results.
- Fix typo in models.py: 'homogenuous' -> 'homogeneous'.

Author: SebastianAment

boxcrete/models.py

@@ -314,7 +314,7 @@ def fit_gwp_gp(
     # GWP is a linear function of the inputs
     covar_module = LinearKernel()
     # removing any input and outcome transforms, as well as the prior mean from
-    # the model to force it to be homogenuous, i.e. it has no offset.
+    # the model to force it to be homogeneous, i.e. it has no offset.
     model_kwargs = {
         "train_X": X,
         "train_Y": Y,