Fix titanic tutorial to convert datatypes to float (#1331)

Summary:
Currently, the Titanic tutorial for Captum leads to an error due to mismatched datatypes within the training data. Explicitly converting to a numpy array with floats resolves this issue.

Pull Request resolved: #1331

Reviewed By: jjuncho

Differential Revision: D61443014

Pulled By: vivekmig

fbshipit-source-id: a7696fa9ce3f14256debe895571e402f35f025c4

Author: vivekmig

tutorials/Titanic_Basic_Interpret.ipynb

@@ -136,9 +136,9 @@
     "# Separate training and test sets using \n",
     "train_indices = np.random.choice(len(labels), int(0.7*len(labels)), replace=False)\n",
     "test_indices = list(set(range(len(labels))) - set(train_indices))\n",
-    "train_features = data[train_indices]\n",
+    "train_features = np.array(data[train_indices], dtype=float)\n",
     "train_labels = labels[train_indices]\n",
-    "test_features = data[test_indices]\n",
+    "test_features = np.array(data[test_indices], dtype=float)\n",
     "test_labels = labels[test_indices]"
    ]
   },