updates for sdk/cli inference examples (#2525)

* updates for sdk/cli inference examples

* deploying from azureml-staging erroring out

cli/foundation-models/system/inference/image-classification/image-classification-batch-endpoint.sh

@@ -5,14 +5,13 @@ set -x
 # the sample scoring file available in the same folder as the above notebook
 
 # script inputs
-registry_name="azureml-staging"
+registry_name="azureml-preview"
 subscription_id="<SUBSCRIPTION_ID>"
 resource_group_name="<RESOURCE_GROUP>"
 workspace_name="<WORKSPACE_NAME>"
 
 # This is the model from system registry that needs to be deployed
 model_name="microsoft-beit-base-patch16-224-pt22k-ft22k"
-
 model_label="latest"
 
 deployment_compute="cpu-cluster"

cli/foundation-models/system/inference/image-classification/image-classification-online-endpoint.sh

@@ -3,15 +3,14 @@ set -x
 # the sample scoring file available in the same folder as the above notebook
 
 # script inputs
-registry_name="azureml-staging"
+registry_name="azureml-preview"
 subscription_id="<SUBSCRIPTION_ID>"
 resource_group_name="<RESOURCE_GROUP>"
 workspace_name="<WORKSPACE_NAME>"
 
 # This is the model from system registry that needs to be deployed
 model_name="microsoft-beit-base-patch16-224-pt22k-ft22k"
-# using the latest version of the model - not working yet
-model_version=2
+model_label="latest"
 
 version=$(date +%s)
 endpoint_name="image-classification-$version"
@@ -42,12 +41,15 @@ workspace_info="--resource-group $resource_group_name --workspace-name $workspac
 
 # 2. Check if the model exists in the registry
 # need to confirm model show command works for registries outside the tenant (aka system registry)
-if ! az ml model show --name $model_name --version $model_version --registry-name $registry_name 
+if ! az ml model show --name $model_name --label $model_label --registry-name $registry_name 
 then
-    echo "Model $model_name:$model_version does not exist in registry $registry_name"
+    echo "Model $model_name:$model_label does not exist in registry $registry_name"
     exit 1
 fi
 
+# get the latest model version
+model_version=$(az ml model show --name $model_name --label $model_label --registry-name $registry_name --query version --output tsv)
+
 # 3. Deploy the model to an endpoint
 # create online endpoint 
 az ml online-endpoint create --name $endpoint_name $workspace_info  || {

cli/foundation-models/system/inference/image-instance-segmentation/image-instance-segmentation-batch-endpoint.sh

@@ -2,14 +2,13 @@ set -x
 # the commands in this file map to steps in this notebook: https://aka.ms/azureml-infer-batch-sdk-image-instance-segmentation
 
 # script inputs
-registry_name="azureml-staging"
+registry_name="azureml-preview"
 subscription_id="<SUBSCRIPTION_ID>"
 resource_group_name="<RESOURCE_GROUP>"
 workspace_name="<WORKSPACE_NAME>"
 
 # This is the model from system registry that needs to be deployed
 model_name="mask_rcnn_swin-t-p4-w7_fpn_1x_coco"
-
 model_label="latest"
 
 deployment_compute="cpu-cluster"

cli/foundation-models/system/inference/image-instance-segmentation/image-instance-segmentation-online-endpoint.sh

@@ -3,14 +3,13 @@ set -x
 # the sample scoring file available in the same folder as the above notebook
 
 # script inputs
-registry_name="azureml-staging"
+registry_name="azureml-preview"
 subscription_id="<SUBSCRIPTION_ID>"
 resource_group_name="<RESOURCE_GROUP>"
 workspace_name="<WORKSPACE_NAME>"
 
 # This is the model from system registry that needs to be deployed
 model_name="mask_rcnn_swin-t-p4-w7_fpn_1x_coco"
-# using the latest version of the model - not working yet
 model_label="latest"
 
 version=$(date +%s)

cli/foundation-models/system/inference/image-object-detection/image-object-detection-batch-endpoint.sh

@@ -2,14 +2,13 @@ set -x
 # the commands in this file map to steps in this notebook: https://aka.ms/azureml-infer-batch-sdk-image-object-detection
 
 # script inputs
-registry_name="azureml-staging"
+registry_name="azureml-preview"
 subscription_id="<SUBSCRIPTION_ID>"
 resource_group_name="<RESOURCE_GROUP>"
 workspace_name="<WORKSPACE_NAME>"
 
 # This is the model from system registry that needs to be deployed
 model_name="yolof_r50_c5_8x8_1x_coco"
-
 model_label="latest"
 
 deployment_compute="cpu-cluster"

cli/foundation-models/system/inference/image-object-detection/image-object-detection-online-endpoint.sh

@@ -3,14 +3,13 @@ set -x
 # the sample scoring file available in the same folder as the above notebook
 
 # script inputs
-registry_name="azureml-staging"
+registry_name="azureml-preview"
 subscription_id="<SUBSCRIPTION_ID>"
 resource_group_name="<RESOURCE_GROUP>"
 workspace_name="<WORKSPACE_NAME>"
 
 # This is the model from system registry that needs to be deployed
 model_name="yolof_r50_c5_8x8_1x_coco"
-# using the latest version of the model - not working yet
 model_label="latest"
 
 version=$(date +%s)

sdk/python/foundation-models/system/inference/image-classification/image-classification-batch-endpoint.ipynb

@@ -16,7 +16,7 @@
     "* MultiLabel: An image can be categorised into more than one class.\n",
     " \n",
     "### Model\n",
-    "Models that can perform the `image-classification` task are tagged with `image-classification`. We will use the `microsoft-beit-base-patch16-224-pt22k-ft22k` model in this notebook. If you opened this notebook from a specific model card, remember to replace the specific model name. If you don't find a model that suits your scenario or domain, you can discover and [import models from HuggingFace hub](../../import/import-model-from-huggingface.ipynb) and then use them for inference. \n",
+    "Models that can perform the `image-classification` task are tagged with `image-classification`. We will use the `microsoft-beit-base-patch16-224-pt22k-ft22k` model in this notebook. If you opened this notebook from a specific model card, remember to replace the specific model name. If you don't find a model that suits your scenario or domain, you can discover and [import models from HuggingFace hub](../../import/import_model_into_registry.ipynb) and then use them for inference. \n",
     "\n",
     "### Inference data\n",
     "We will use the [fridgeObjects](https://cvbp-secondary.z19.web.core.windows.net/datasets/image_classification/fridgeObjects.zip) dataset.\n",
@@ -76,12 +76,12 @@
     "\n",
     "workspace_ml_client = MLClient(credential, subscription_id, resource_group, workspace_name)\n",
     "\n",
-    "# the models, fine tuning pipelines and environments are available in the AzureML system registry, \"azureml-staging\"\n",
+    "# the models, fine tuning pipelines and environments are available in the AzureML system registry, \"azureml-preview\"\n",
     "registry_ml_client = MLClient(\n",
     "    credential,\n",
     "    subscription_id,\n",
     "    resource_group,\n",
-    "    registry_name=\"azureml-staging\",\n",
+    "    registry_name=\"azureml-preview\",\n",
     ")\n",
     "# generating a unique timestamp that can be used for names and versions that need to be unique\n",
     "timestamp = str(int(time.time()))"
@@ -92,7 +92,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "#### Create a compute cluster.\n",
+    "#### Create a compute cluster\n",
     "Use the model card from the AzureML system registry to check the minimum required inferencing SKU, referenced as size below. If you already have a sufficient compute cluster, you can simply define the name in compute_name in the following code block."
    ]
   },
@@ -130,7 +130,7 @@
    "source": [
     "### 2. Pick a model to deploy\n",
     "\n",
-    "Browse models in the Model Catalog in the AzureML Studio, filtering by the `image-classification` task. In this example, we use the `microsoft-beit-base-patch16-224-pt22k-ft22k ` model. If you have opened this notebook for a different model, replace the model name and version accordingly. "
+    "Browse models in the Model Catalog in the AzureML Studio, filtering by the `image-classification` task. In this example, we use the `microsoft-beit-base-patch16-224-pt22k-ft22k ` model. If you have opened this notebook for a different model, replace the model name accordingly. "
    ]
   },
   {
@@ -239,7 +239,7 @@
     "\n",
     "We can provide input images to batch inference either in a folder containing images or in a csv file containing \"image\" named column having images in base 64 format.\n",
     "\n",
-    "Note: If job failed with error Assertion Error (The actual length exceeded max length 100 MB) then please try with less number of input images or use ImageFolder Input mode."
+    "Note: If job failed with error Assertion Error (`The actual length exceeded max length 100 MB`) then please try with less number of input images or use ImageFolder Input mode."
    ]
   },
   {
@@ -276,12 +276,6 @@
     "Image(filename=sample_image)"
    ]
   },
-  {
-   "attachments": {},
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": []
-  },
   {
    "attachments": {},
    "cell_type": "markdown",

sdk/python/foundation-models/system/inference/image-classification/image-classification-online-endpoint.ipynb

@@ -16,7 +16,7 @@
     "* MultiLabel: An image can be categorised into more than one class.\n",
     " \n",
     "### Model\n",
-    "Models that can perform the `image-classification` task are tagged with `image-classification`. We will use the `microsoft-beit-base-patch16-224-pt22k-ft22k` model in this notebook. If you opened this notebook from a specific model card, remember to replace the specific model name. If you don't find a model that suits your scenario or domain, you can discover and [import models from HuggingFace hub](../../import/import-model-from-huggingface.ipynb) and then use them for inference. \n",
+    "Models that can perform the `image-classification` task are tagged with `image-classification`. We will use the `microsoft-beit-base-patch16-224-pt22k-ft22k` model in this notebook. If you opened this notebook from a specific model card, remember to replace the specific model name. If you don't find a model that suits your scenario or domain, you can discover and [import models from HuggingFace hub](../../import/import_model_into_registry.ipynb) and then use them for inference.\n",
     "\n",
     "### Inference data\n",
     "We will use the [fridgeObjects](https://cvbp-secondary.z19.web.core.windows.net/datasets/image_classification/fridgeObjects.zip) dataset.\n",
@@ -39,7 +39,7 @@
     "### 1. Setup pre-requisites\n",
     "* Install dependencies\n",
     "* Connect to AzureML Workspace. Learn more at [set up SDK authentication](https://learn.microsoft.com/en-us/azure/machine-learning/how-to-setup-authentication?tabs=sdk). Replace  `<WORKSPACE_NAME>`, `<RESOURCE_GROUP>` and `<SUBSCRIPTION_ID>` below.\n",
-    "* Connect to `azureml-staging` system registry"
+    "* Connect to `azureml-preview` system registry"
    ]
   },
   {
@@ -72,13 +72,12 @@
     "    workspace_name = \"<AML_WORKSPACE_NAME>\"\n",
     "workspace_ml_client = MLClient(credential, subscription_id, resource_group, workspace_name)\n",
     "\n",
-    "# the models, fine tuning pipelines and environments are available in the AzureML system registry, \"azureml-staging\"\n",
+    "# the models, fine tuning pipelines and environments are available in the AzureML system registry, \"azureml-preview\"\n",
     "registry_ml_client = MLClient(\n",
     "    credential,\n",
     "    subscription_id,\n",
     "    resource_group,\n",
-    "    # workspace_name\n",
-    "    registry_name=\"azureml-staging\",\n",
+    "    registry_name=\"azureml-preview\",\n",
     ")\n",
     "# genrating a unique timestamp that can be used for names and versions that need to be unique\n",
     "timestamp = str(int(time.time()))"
@@ -91,7 +90,7 @@
    "source": [
     "### 2. Pick a model to deploy\n",
     "\n",
-    "Browse models in the Model Catalog in the AzureML Studio, filtering by the `image-classification` task. In this example, we use the `microsoft-beit-base-patch16-224-pt22k-ft22k ` model. If you have opened this notebook for a different model, replace the model name and version accordingly. "
+    "Browse models in the Model Catalog in the AzureML Studio, filtering by the `image-classification` task. In this example, we use the `microsoft-beit-base-patch16-224-pt22k-ft22k ` model. If you have opened this notebook for a different model, replace the model name accordingly. "
    ]
   },
   {
@@ -101,9 +100,8 @@
    "outputs": [],
    "source": [
     "model_name = \"microsoft-beit-base-patch16-224-pt22k-ft22k\"\n",
-    "model_version = \"1\"\n",
-    "foundation_model = registry_ml_client.models.get(model_name, model_version)\n",
-    "\n",
+    "foundation_models = registry_ml_client.models.list(name=model_name)\n",
+    "foundation_model = max(foundation_models, key=lambda x: x.version)\n",
     "print(\n",
     "    f\"\\n\\nUsing model name: {foundation_model.name}, version: {foundation_model.version}, id: {foundation_model.id} for inferencing\"\n",
     ")"
@@ -226,7 +224,7 @@
     "    endpoint_name=online_endpoint_name,\n",
     "    model=foundation_model.id,\n",
     "    # use GPU instance type like Standard_NC6s_v3 for faster explanations\n",
-    "    instance_type=\"Standard_DS3_V2\",  # \"Standard_DS3_V2\",\n",
+    "    instance_type=\"Standard_DS3_V2\",\n",
     "    instance_count=1,\n",
     "    request_settings=OnlineRequestSettings(\n",
     "        max_concurrent_requests_per_instance=1, request_timeout_ms=5000, max_queue_wait_ms=500  # 90000,\n",
@@ -258,7 +256,7 @@
    "source": [
     "### 5. Test the endpoint with sample data\n",
     "\n",
-    "We will fetch some sample data from the test dataset and submit to online endpoint for inference. We will then show the display the scored labels alongside the ground truth labels"
+    "We will fetch some sample data from the test dataset and submit to online endpoint for inference. We will then show the scored labels alongside the ground truth labels."
    ]
   },
   {

sdk/python/foundation-models/system/inference/image-instance-segmentation/image-instance-segmentation-batch-endpoint.ipynb

@@ -73,12 +73,12 @@
     "\n",
     "workspace_ml_client = MLClient(credential, subscription_id, resource_group, workspace_name)\n",
     "\n",
-    "# the models, fine tuning pipelines and environments are available in the AzureML system registry, \"azureml-staging\"\n",
+    "# the models, fine tuning pipelines and environments are available in the AzureML system registry, \"azureml-preview\"\n",
     "registry_ml_client = MLClient(\n",
     "    credential,\n",
     "    subscription_id,\n",
     "    resource_group,\n",
-    "    registry_name=\"azureml-staging\",\n",
+    "    registry_name=\"azureml-preview\",\n",
     ")\n",
     "# generating a unique timestamp that can be used for names and versions that need to be unique\n",
     "timestamp = str(int(time.time()))"
@@ -127,7 +127,7 @@
    "source": [
     "### 2. Pick a model to deploy\n",
     "\n",
-    "Browse models in the Model Catalog in the AzureML Studio, filtering by the `image-instance-segmentation` task. In this example, we use the `mask_rcnn_swin-t-p4-w7_fpn_1x_coco` model. If you have opened this notebook for a different model, replace the model name and version accordingly. This is a pre-trained model and may not give correct prediction for your dataset. You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference. Please refer to the [notebook.](../../finetune/image-instance-segmentation/mmdetection-fridgeobjects-instance-segmentation.ipynb)"
+    "Browse models in the Model Catalog in the AzureML Studio, filtering by the `image-instance-segmentation` task. In this example, we use the `mask_rcnn_swin-t-p4-w7_fpn_1x_coco` model. If you have opened this notebook for a different model, replace the model name accordingly. This is a pre-trained model and may not give correct prediction for your dataset. We strongly recommend to TRAIN this model on a down-stream task to be able to use it for predictions and inference. Please refer to the [notebook.](../../finetune/image-instance-segmentation/mmdetection-fridgeobjects-instance-segmentation.ipynb)"
    ]
   },
   {
@@ -137,9 +137,8 @@
    "outputs": [],
    "source": [
     "model_name = \"mask_rcnn_swin-t-p4-w7_fpn_1x_coco\"\n",
-    "\n",
-    "foundation_model = registry_ml_client.models.get(model_name, label=\"latest\")\n",
-    "\n",
+    "foundation_models = registry_ml_client.models.list(name=model_name)\n",
+    "foundation_model = max(foundation_models, key=lambda x: x.version)\n",
     "print(\n",
     "    f\"\\n\\nUsing model name: {foundation_model.name}, version: {foundation_model.version}, id: {foundation_model.id} for inferencing\"\n",
     ")"
@@ -481,11 +480,6 @@
   }
  ],
  "metadata": {
-  "kernelspec": {
-   "display_name": "temp",
-   "language": "python",
-   "name": "python3"
-  },
   "language_info": {
    "name": "python"
   }

sdk/python/foundation-models/system/inference/image-instance-segmentation/image-instance-segmentation-online-endpoint.ipynb

@@ -38,7 +38,7 @@
     "### 1. Setup pre-requisites\n",
     "* Install dependencies\n",
     "* Connect to AzureML Workspace. Learn more at [set up SDK authentication](https://learn.microsoft.com/en-us/azure/machine-learning/how-to-setup-authentication?tabs=sdk). Replace  `<WORKSPACE_NAME>`, `<RESOURCE_GROUP>` and `<SUBSCRIPTION_ID>` below.\n",
-    "* Connect to `azureml` system registry"
+    "* Connect to `azureml-preview` system registry"
    ]
   },
   {
@@ -71,13 +71,12 @@
     "    workspace_name = \"<AML_WORKSPACE_NAME>\"\n",
     "workspace_ml_client = MLClient(credential, subscription_id, resource_group, workspace_name)\n",
     "\n",
-    "# the models, fine tuning pipelines and environments are available in the AzureML system registry, \"azureml-staging\"\n",
+    "# the models, fine tuning pipelines and environments are available in the AzureML system registry, \"azureml-preview\"\n",
     "registry_ml_client = MLClient(\n",
     "    credential,\n",
     "    subscription_id,\n",
     "    resource_group,\n",
-    "    # workspace_name\n",
-    "    registry_name=\"azureml-staging\",\n",
+    "    registry_name=\"azureml-preview\",\n",
     ")\n",
     "# genrating a unique timestamp that can be used for names and versions that need to be unique\n",
     "timestamp = str(int(time.time()))"
@@ -90,7 +89,7 @@
    "source": [
     "### 2. Pick a model to deploy\n",
     "\n",
-    "Browse models in the Model Catalog in the AzureML Studio, filtering by the `image-instance-segmentation` task. In this example, we use the `mask_rcnn_swin-t-p4-w7_fpn_1x_coco` model. If you have opened this notebook for a different model, replace the model name and version accordingly. This is a pre-trained model and may not give correct prediction for your dataset. You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference. Please refer to the [notebook.](../../finetune/image-instance-segmentation/mmdetection-fridgeobjects-instance-segmentation.ipynb)"
+    "Browse models in the Model Catalog in the AzureML Studio, filtering by the `image-instance-segmentation` task. In this example, we use the `mask_rcnn_swin-t-p4-w7_fpn_1x_coco` model. If you have opened this notebook for a different model, replace the model name accordingly. This is a pre-trained model and may not give correct prediction for your dataset. We strongly recommend to TRAIN this model on a down-stream task to be able to use it for predictions and inference. Please refer to the [notebook](../../finetune/image-instance-segmentation/mmdetection-fridgeobjects-instance-segmentation.ipynb)."
    ]
   },
   {
@@ -100,9 +99,8 @@
    "outputs": [],
    "source": [
     "model_name = \"mask_rcnn_swin-t-p4-w7_fpn_1x_coco\"\n",
-    "# foundation_model = registry_ml_client.models.get(model_name, label=\"latest\")\n",
-    "foundation_model = workspace_ml_client.models.get(model_name, label=\"latest\")\n",
-    "\n",
+    "foundation_models = registry_ml_client.models.list(name=model_name)\n",
+    "foundation_model = max(foundation_models, key=lambda x: x.version)\n",
     "print(\n",
     "    f\"\\n\\nUsing model name: {foundation_model.name}, version: {foundation_model.version}, id: {foundation_model.id} for inferencing\"\n",
     ")"
@@ -221,7 +219,9 @@
     "    instance_type=\"Standard_DS3_v2\",\n",
     "    instance_count=1,\n",
     "    request_settings=OnlineRequestSettings(\n",
-    "        max_concurrent_requests_per_instance=1, request_timeout_ms=5000, max_queue_wait_ms=500  # 90000,\n",
+    "        max_concurrent_requests_per_instance=1,\n",
+    "        request_timeout_ms=5000, # 90000,\n",
+    "        max_queue_wait_ms=500\n",
     "    ),\n",
     "    liveness_probe=ProbeSettings(\n",
     "        failure_threshold=30,\n",
@@ -286,13 +286,10 @@
     "\n",
     "sample_image = os.path.join(dataset_dir, \"images\", \"99.jpg\")\n",
     "\n",
-    "\n",
     "def read_image(image_path):\n",
     "    with open(image_path, \"rb\") as f:\n",
     "        return f.read()\n",
     "\n",
-    "\n",
-    "# {\"inputs\":{\"image\":[\"\"]}}\n",
     "request_json = {\n",
     "    \"input_data\": \n",
     "        {\n",
@@ -350,11 +347,6 @@
   }
  ],
  "metadata": {
-  "kernelspec": {
-   "display_name": "temp",
-   "language": "python",
-   "name": "python3"
-  },
   "language_info": {
    "name": "python"
   }