Updated to AI Platform

examples/cloudml-bank-marketing/bank_marketing_classification_model.ipynb

@@ -1328,7 +1328,7 @@
     "id": "cAfB1M0KzAdP"
    },
    "source": [
-    "Let us define some environment variables that CMLE uses. Do not forget to replace all the variables in square brackets (along with the square brackets) with your credentials."
+    "Let us define some environment variables that AI Platform uses. Do not forget to replace all the variables in square brackets (along with the square brackets) with your credentials."
    ]
   },
   {
@@ -1342,7 +1342,6 @@
    "outputs": [],
    "source": [
     "%env GCS_BUCKET=<GCS_BUCKET>\n",
-    "%env PROJECT_ID=<PROJECT_ID> \n",
     "%env REGION=us-central1\n",
     "%env LOCAL_DIRECTORY=./trainer/data\n",
     "%env TRAINER_PACKAGE_PATH=./trainer"
@@ -1355,7 +1354,7 @@
     "id": "QWQ8kPKc9rYk"
    },
    "source": [
-    "ML Engine needs a Python package with our code to train models. We need to create a directory and move our code there. We also need to create an `__init__.py` file, this is a unique feature of python. You can read the [docs](https://docs.python.org/3/tutorial/modules.html#packages) to understand more about this file."
+    "AI Platform needs a Python package with our code to train models. We need to create a directory and move our code there. We also need to create an `__init__.py` file, this is a unique feature of python. You can read the [docs](https://docs.python.org/3/tutorial/modules.html#packages) to understand more about this file."
    ]
   },
   {
@@ -1590,7 +1589,7 @@
     "id": "yStJtBfqFlGF"
    },
    "source": [
-    "To actully run the `train.py` file we need some parameters so that CMLE knows how to set up the environment to run sucessfully."
+    "To actully run the `train.py` file we need some parameters so that AI Platform knows how to set up the environment to run sucessfully."
    ]
   },
   {
@@ -1609,7 +1608,7 @@
     "\n",
     "echo $JOBNAME \n",
     "\n",
-    "gcloud ml-engine jobs submit training model_training_$JOBNAME \\\n",
+    "gcloud ai-platform jobs submit training model_training_$JOBNAME \\\n",
     "        --job-dir $GCS_BUCKET/$JOBNAME/output \\\n",
     "        --package-path trainer \\\n",
     "        --module-name trainer.task \\\n",
@@ -1626,7 +1625,7 @@
     "id": "EdG8UV2RL5tV"
    },
    "source": [
-    "The above cell submits a job to CMLE which you can view by going to the Google Cloud Console's sidebar and select `ML Engine > Jobs` or search `ML Engine` in the search bar. ONLY run the cells below after your job has completed sucessfully. (It should take approximately 8 minutes to run).\n",
+    "The above cell submits a job to AI Platform which you can view by going to the Google Cloud Console's sidebar and select `AI Platform > Jobs` or search `AI Platform` in the search bar. ONLY run the cells below after your job has completed sucessfully. (It should take approximately 8 minutes to run).\n",
     "\n"
    ]
   },
@@ -1672,18 +1671,40 @@
     "test = test_prepared.as_matrix().tolist()"
    ]
   },
+  {
+    "cell_type": "markdown",
+    "metadata": {
+     "colab_type": "text",
+     "id": "oWGm6c6KIs2O"
+    },
+    "source": [
+     "Next, find the model directory in your GCS bucket that contains the model created in the previous steps."
+    ]
+   },
+   {
+    "cell_type": "code",
+    "execution_count": 88,
+    "metadata": {
+     "colab": {},
+     "colab_type": "code",
+     "id": "tJdbJg2jNd05"
+    },
+    "source": [
+     "!gsutil ls $GCS_BUCKET"
+    ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {
     "colab_type": "text",
     "id": "8AlU8Bo2KJS8"
    },
    "source": [
-    "Just like training in CMLE, we set some environment variables when we run our command line commands. Note that `<MODEL_DIRECTORY>` can be found in your bucket, it will be of the form `model_YYYYMMDD_HHMMSS` (e.g. `model_190114_134228`)."
+    "Just like training in AI Platform, we set some environment variables when we run our command line commands. Note that `<GCS_BUCKET>` is your the name of your GCS bucket set earlier. The `MODEL_DIRECTORY` will be inside the GCS bucket and of the form `model_YYYYMMDD_HHMMSS` (e.g. `model_190114_134228`)."
    ]
   },
   {
-   "cell_type": "code",
+   "cell_type": "code",	
    "execution_count": null,
    "metadata": {
     "colab": {},
@@ -1695,7 +1716,7 @@
     "%env VERSION_NAME=v1\n",
     "%env MODEL_NAME=cmle_model\n",
     "%env JSON_INSTANCE=input.json\n",
-    "%env MODEL_DIR=gs://<MODEL_DIRECTORY>/model_YYYYMMDD_HHMMSS\n",
+    "%env MODEL_DIR=gs://<GCS_BUCKET>/MODEL_DIRECTORY\n",
     "%env FRAMEWORK=SCIKIT_LEARN"
    ]
   },
@@ -1719,7 +1740,7 @@
    },
    "outputs": [],
    "source": [
-    "! gcloud ml-engine models create $MODEL_NAME --regions=us-central1"
+    "! gcloud ai-platform models create $MODEL_NAME --regions=us-central1"
    ]
   },
   {
@@ -1732,7 +1753,7 @@
    },
    "outputs": [],
    "source": [
-    "! gcloud ml-engine versions create $VERSION_NAME \\\n",
+    "! gcloud ai-platform versions create $VERSION_NAME \\\n",
     "        --model $MODEL_NAME --origin $MODEL_DIR \\\n",
     "        --runtime-version 1.9 --framework $FRAMEWORK \\\n",
     "        --python-version 3.5"
@@ -1796,7 +1817,7 @@
    },
    "outputs": [],
    "source": [
-    "! gcloud ml-engine predict --model $MODEL_NAME \\\n",
+    "! gcloud ai-platform predict --model $MODEL_NAME \\\n",
     "        --version $VERSION_NAME \\\n",
     "        --json-instances $JSON_INSTANCE"
    ]
@@ -1825,11 +1846,11 @@
     "import os\n",
     "import pandas as pd\n",
     "\n",
-    "PROJECT_ID = os.environ['PROJECT_ID']\n",
+    "PROJECT_ID = os.environ['GOOGLE_CLOUD_PROJECT']\n",
     "VERSION_NAME = os.environ['VERSION_NAME']\n",
     "MODEL_NAME = os.environ['MODEL_NAME']\n",
     "\n",
-    "# Create our CMLE service\n",
+    "# Create our AI Platform service\n",
     "service = googleapiclient.discovery.build('ml', 'v1')\n",
     "name = 'projects/{}/models/{}'.format(PROJECT_ID, MODEL_NAME)\n",
     "name += '/versions/{}'.format(VERSION_NAME)\n",