Update docs

Author: ospillinger

cortex-installer.sh

@@ -165,9 +165,10 @@ function install_operator() {
   setup_bucket
   setup_cloudwatch_logs
 
+  prompt_for_telemetry
+
   echo "Installing the Cortex operator ..."
 
-  prompt_for_telemetry
   setup_namespace
   setup_configmap
   setup_secrets
@@ -1933,7 +1934,7 @@ function prompt_for_telemetry() {
     while true
     do
       echo
-      read -p "Would you like to help improve Cortex by anonymously sending error reports and usage stats to the dev team [Y/n] " -n 1 -r
+      read -p "Would you like to help improve Cortex by anonymously sending error reports and usage stats to the dev team? [Y/n] " -n 1 -r
       echo
       if [[ $REPLY =~ ^[Yy]$ ]]; then
         export CORTEX_ENABLE_TELEMETRY=true

docs/operator/install.md

@@ -42,7 +42,7 @@ Cortex runs on Kubernetes. If you don't already have a Kubernetes cluster, [eksc
 eksctl create cluster --name=cortex --nodes=2 --node-type=t3.medium
 ```
 
-This cluster configuration will cost about $0.29 per hour.
+This cluster configuration will cost about $0.29 per hour in AWS fees.
 
 ## Operator
 

docs/quick-start.md

@@ -270,6 +270,8 @@ Add to `app.yaml`:
   training:
     num_steps: 1000
     batch_size: 10
+  aggregates:
+    - class_index
 ```
 
 #### Implement the estimator
@@ -292,7 +294,7 @@ def create_estimator(run_config, model_config):
     return tf.estimator.DNNClassifier(
         feature_columns=feature_columns,
         hidden_units=model_config["hparams"]["hidden_units"],
-        n_classes=3,
+        n_classes=len(model_config["aggregates"]["class_index"]),
         config=run_config,
     )
 ```
@@ -330,6 +332,12 @@ You can get an overview of the deployment using `cortex get` (see [resource stat
 ```
 $ cortex get
 
+---------------
+Python Packages
+---------------
+
+None
+
 -----------
 Raw Columns
 -----------
@@ -394,6 +402,7 @@ You can get a summary of the status of resources using `cortex status`:
 ```
 $ cortex status --watch
 
+Python Packages:       none
 Raw Columns:           5 ready
 Aggregates:            9 ready
 Transformed Columns:   5 ready

examples/iris/resources/models.yaml

@@ -1,5 +1,6 @@
 - kind: model
   name: dnn
+  path: dnn.py
   type: classification
   target_column: class_indexed
   feature_columns:
@@ -10,8 +11,8 @@
   hparams:
     hidden_units: [4, 2]
   data_partition_ratio:
-    training: 0.9
-    evaluation: 0.1
+    training: 80
+    evaluation: 20
   training:
     batch_size: 10
     num_steps: 1000