diff --git a/source/notebooks/sagemaker_fraud_detection.ipynb b/source/notebooks/sagemaker_fraud_detection.ipynb
index 8530358..2f50ddf 100644
--- a/source/notebooks/sagemaker_fraud_detection.ipynb
+++ b/source/notebooks/sagemaker_fraud_detection.ipynb
@@ -310,12 +310,9 @@
    },
    "outputs": [],
    "source": [
-    "from sagemaker.serializers import CSVSerializer\n",
+    "from sagemaker.serializers import CSVSerializer \n",
     "from sagemaker.deserializers import JSONDeserializer\n",
-    " \n",
-    "rcf_predictor.content_type = 'text/csv'\n",
-    "rcf_predictor.serializer = CSVSerializer()\n",
-    "rcf_predictor.accept = 'application/json'\n",
+    "rcf_predictor.serializer = CSVSerializer() \n",
     "rcf_predictor.deserializer = JSONDeserializer()"
    ]
   },
@@ -372,6 +369,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
+    "%pip install seaborn --q\n",
     "import seaborn as sns\n",
     "import matplotlib.pyplot as plt\n",
     "sns.set(color_codes=True)"
@@ -655,12 +653,12 @@
    "source": [
     "from sagemaker.serializers import CSVSerializer\n",
     "\n",
+    "\n",
     "predictor = clf.deploy(initial_instance_count=1,\n",
     "                       model_name=\"{}-xgb\".format(config.SOLUTION_PREFIX),\n",
     "                       endpoint_name=\"{}-xgb\".format(config.SOLUTION_PREFIX),\n",
     "                       instance_type=instance_type,\n",
-    "                       serializer=CSVSerializer(),\n",
-    "                       deserializer=None)"
+    "                       serializer=CSVSerializer())"
    ]
   },
   {
@@ -696,7 +694,7 @@
     "    split_array = np.array_split(data, int(data.shape[0] / float(rows) + 1))\n",
     "    predictions = ''\n",
     "    for array in split_array:\n",
-    "        predictions = ','.join([predictions, current_predictor.predict(array).decode('utf-8')])\n",
+    "        predictions = ','.join([predictions, current_predictor.predict(array)])\n",
     "\n",
     "    return np.fromstring(predictions[1:], sep=',')"
    ]
@@ -868,6 +866,18 @@
     "pip install aws_requests_auth"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import logging\n",
+    "\n",
+    "for name in (\"botocore.credentials\", \"botocore\", \"boto3\", \"s3transfer\", \"urllib3\"):\n",
+    "    logging.getLogger(name).setLevel(logging.WARNING)   # or logging.ERROR / logging.CRITICAL"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -1065,13 +1075,25 @@
     "# Specify input and output formats.\n",
     "smote_predictor.content_type = 'text/csv'\n",
     "csv_serializer = CSVSerializer()\n",
-    "smote_predictor.serializer = csv_serializer\n",
     "\n",
     "# Set the deserializer to handle the response from the inference endpoint\n",
     "#csv_deserializer = CSVDeserializer()\n",
     "#smote_predictor.deserializer = csv_deserializer"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from sagemaker.serializers import CSVSerializer\n",
+    "from sagemaker.deserializers import StringDeserializer  # predictions often come back as text\n",
+    "\n",
+    "smote_predictor.serializer = CSVSerializer()            # sets Content-Type to text/csv\n",
+    "smote_predictor.deserializer = StringDeserializer()"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,