Update deployment docs

docs/DetectionSolutionAccelerator/Model-Deployment-Cloud.md

@@ -19,10 +19,18 @@ When deploying a model there are several parameters to be set. In order to make
     "ENV_CONFIG_FILE": "dev_config.json",
     "EXPERIMENT" : "",
     "RUN_ID" : "",
-    "USE_ACI" : true,
+    "TF_VERSION": 2,
+    "REG_MODEL" : true,
     "IMAGE_TYPE": "",
-    "COMPUTE_TARGET_NAME" : "",
-    "REG_MODEL" : false
+    "ACI_PARAMS": {
+        "USE_ACI": true,
+        "ACI_AUTH": true
+    },
+    "AKS_PARAMS": {
+        "USE_AKS": false,
+        "VM_TYPE": "Standard_NC6",
+        "COMPUTE_TARGET_NAME" : ""
+    }
 }
 ```
 
@@ -34,13 +42,23 @@ Below a breakdown of each parameter can be obtained:
 
 - `RUN_ID` - This is the run ID from AML of the model you wish to deploy. This can be found from the Experiment view in AML
 
-- `USE_ACI` - Set this to true if you wish to deploy a model to ACI (this is for testing purposes only)
+- `TF_VERSION` - This is the tensorflow version (1 or 2), needed to specify different scoring script
+
+- `REG_MODEL` - If the model needs to be registered or not from the run_id. If set to false the deployment will check the model registry for a registered model with a runid tag matching that of your set RUN_ID
 
 - `IMAGE_TYPE` - The type of image or "use case" name as defined in the storage as the use case parent folder.
 
-- `COMPUTE_TARGET_NAME` - This will be the name of your deployed service and compute. If this already exists it will be updated
+- `ACI_PARAMS.USE_ACI` - Set this to true if you wish to deploy a model to ACI (this is for testing purposes only)
+
+- `ACI_PARAMS.ACI_AUTH` - Whether to enable key-based authentication when using ACI, default value is true
+
+- `AKS_PARAMS.USE_AKS` - Set this to true if you wish to deploy a model to AKS, this is more suitable for production usage
+
+- `AKS_PARAMS.COMPUTE_TARGET_NAME` - This will be the name of your deployed service and compute. If this already exists it will be updated
+
+- `AKS_PARAMS.VM_TYPE` - Machine type of AKS cluster, default value is "Standard_NC6"
+
 
-- `REG_MODEL` - if the model needs to be registered or not from the run_id. If set to false the deployment will check the model registry for a registered model with a runid tag matching that of your set RUN_ID
 
 ## Running Model Deployment
 

docs/DetectionSolutionAccelerator/Quickstart.md

@@ -213,7 +213,7 @@ The experiment config stores all the model training run paramters. It needs to b
 
 #### Deployment Config
 
-The deployment config stores the deployment paramters and points to the desired training run to deploy the model from. This config needs to be created under the deployment directory within src called deploy_config.json. An example config is provided already in this location called deploy_config_sample.json. A more detailed breakdown of each paramter can be found under model deployment cloud.
+The deployment config stores the deployment parameters and points to the desired training run to deploy the model from. This config needs to be created under the deployment directory within src called deploy_config.json. An example config is provided already in this location called deploy_config_sample.json. A more detailed breakdown of each paramter can be found under model deployment cloud.
 
 ```json
 {
@@ -221,10 +221,17 @@ The deployment config stores the deployment paramters and points to the desired
     "EXPERIMENT" : "",
     "RUN_ID" : "",
     "TF_VERSION": 2,
-    "USE_ACI" : true,
+    "REG_MODEL" : true,
     "IMAGE_TYPE": "",
-    "COMPUTE_TARGET_NAME" : "",
-    "REG_MODEL" : false
+    "ACI_PARAMS": {
+        "USE_ACI": true,
+        "ACI_AUTH": true
+    },
+    "AKS_PARAMS": {
+        "USE_AKS": false,
+        "VM_TYPE": "Standard_NC6",
+        "COMPUTE_TARGET_NAME" : ""
+    }
 }
 ```