You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: articles/machine-learning/tutorial-enable-materialization-backfill-data.md
+3-4Lines changed: 3 additions & 4 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -17,22 +17,21 @@ ms.custom: sdkv2, build-2023
17
17
18
18
# Tutorial 2: Enable materialization and backfill feature data (preview)
19
19
20
-
This tutorial series shows how features seamlessly integrate all phases of the machine learning (ML) lifecycle: prototyping, training, and operationalization.
20
+
This tutorial series shows how features seamlessly integrate all phases of the machine learning lifecycle: prototyping, training, and operationalization.
21
21
22
22
This tutorial is the second part of a four-part series. The first tutorial showed how to create a feature set specification with custom transformations, and then use that feature set to generate training data. This tutorial describes materialization.
23
23
24
24
Materialization computes the feature values for a feature window and then stores those values in a materialization store. All feature queries can then use the values from the materialization store.
25
25
26
-
A feature set query applies the transformations to the source on the fly, to compute the features before it returns the values. This process works well for the prototyping phase. However, for training and inference operations in a production environment, we recommend that you materialize the features for greater reliability and availability.
26
+
Without materialization, a feature set query applies the transformations to the source on the fly, to compute the features before it returns the values. This process works well for the prototyping phase. However, for training and inference operations in a production environment, we recommend that you materialize the features for greater reliability and availability.
27
27
28
28
In this tutorial, you learn how to:
29
29
30
30
> [!div class="checklist"]
31
31
> * Enable an offline store on the feature store by creating and attaching an Azure Data Lake Storage Gen2 container and a user-assigned managed identity (UAI).
32
32
> * Enable offline materialization on the feature sets, and backfill the feature data.
33
33
34
-
> [!IMPORTANT]
35
-
> This feature is currently in public preview. This preview version is provided without a service-level agreement, and we don't recommend it for production workloads. Certain features might not be supported or might have constrained capabilities. For more information, see [Supplemental Terms of Use for Microsoft Azure Previews](https://azure.microsoft.com/support/legal/preview-supplemental-terms/).
Copy file name to clipboardExpand all lines: articles/machine-learning/tutorial-enable-recurrent-materialization-run-batch-inference.md
+2-3Lines changed: 2 additions & 3 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -17,7 +17,7 @@ ms.custom: sdkv2, build-2023
17
17
18
18
# Tutorial 4: Enable recurrent materialization and run batch inference (preview)
19
19
20
-
This tutorial series shows how features seamlessly integrate all phases of the machine learning (ML) lifecycle: prototyping, training, and operationalization.
20
+
This tutorial series shows how features seamlessly integrate all phases of the machine learning lifecycle: prototyping, training, and operationalization.
21
21
22
22
The first tutorial showed how to create a feature set specification with custom transformations, and then use that feature set to generate training data. The second tutorial showed how to enable materialization and perform a backfill. The third tutorial showed how to experiment with features as a way to improve model performance. It also showed how a feature store increases agility in the experimentation and training flows.
23
23
@@ -28,8 +28,7 @@ This tutorial explains how to:
28
28
> * Enable recurrent materialization for the `transactions` feature set.
29
29
> * Run a batch inference pipeline on the registered model.
30
30
31
-
> [!IMPORTANT]
32
-
> This feature is currently in public preview. This preview version is provided without a service-level agreement, and we don't recommend it for production workloads. Certain features might not be supported or might have constrained capabilities. For more information, see [Supplemental Terms of Use for Microsoft Azure Previews](https://azure.microsoft.com/support/legal/preview-supplemental-terms/).
Copy file name to clipboardExpand all lines: articles/machine-learning/tutorial-experiment-train-models-using-features.md
+5-6Lines changed: 5 additions & 6 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -17,7 +17,7 @@ ms.custom: sdkv2, build-2023
17
17
18
18
# Tutorial 3: Experiment and train models by using features (preview)
19
19
20
-
This tutorial series shows how features seamlessly integrate all phases of the machine learning (ML) lifecycle: prototyping, training, and operationalization.
20
+
This tutorial series shows how features seamlessly integrate all phases of the machine learning lifecycle: prototyping, training, and operationalization.
21
21
22
22
The first tutorial showed how to create a feature set specification with custom transformations, and then use that feature set to generate training data. The second tutorial showed how to enable materialization and perform a backfill.
23
23
@@ -30,8 +30,7 @@ In this tutorial, you learn how to:
30
30
> * Select features for the model from the `transactions` and `accounts` feature sets, and save them as a feature retrieval specification.
31
31
> * Run a training pipeline that uses the feature retrieval specification to train a new model. This pipeline uses the built-in feature retrieval component to generate the training data.
32
32
33
-
> [!IMPORTANT]
34
-
> This feature is currently in public preview. This preview version is provided without a service-level agreement, and we don't recommend it for production workloads. Certain features might not be supported or might have constrained capabilities. For more information, see [Supplemental Terms of Use for Microsoft Azure Previews](https://azure.microsoft.com/support/legal/preview-supplemental-terms/).
@@ -149,7 +148,7 @@ You don't need to connect to a feature store. In this procedure, you create the
149
148
150
149
## Locally experiment with unregistered features
151
150
152
-
As you develop features, you might want to locally test and validate them before you register them with the feature store or run training pipelines in the cloud. A combination of a local unregistered feature set (`accounts`) and a feature set registered in the feature store (`transactions`) generates training data for the ML model.
151
+
As you develop features, you might want to locally test and validate them before you register them with the feature store or run training pipelines in the cloud. A combination of a local unregistered feature set (`accounts`) and a feature set registered in the feature store (`transactions`) generates training data for the machine learning model.
153
152
154
153
1. Select features for the model.
155
154
@@ -191,15 +190,15 @@ In the following steps, you select a list of features, run a training pipeline,
191
190
192
191
1. Select features for the model, and export the model as a feature retrieval specification.
193
192
194
-
In the previous steps, you selected features from a combination of registered and unregistered feature sets, for local experimentation and testing. You can now experiment in the cloud. Your model-shipping agility increases if you save the selected features as a feature retrieval specification, and then use the specification in the ML operations (MLOps) or continuous integration and continuous delivery (CI/CD) flow for training and inference.
193
+
In the previous steps, you selected features from a combination of registered and unregistered feature sets, for local experimentation and testing. You can now experiment in the cloud. Your model-shipping agility increases if you save the selected features as a feature retrieval specification, and then use the specification in the machine learning operations (MLOps) or continuous integration and continuous delivery (CI/CD) flow for training and inference.
195
194
196
195
1. Select features for the model.
197
196
198
197
[!notebook-python[] (~/azureml-examples-main/sdk/python/featurestore_sample/notebooks/sdk_only/3. Experiment and train models using features.ipynb?name=select-reg-features)]
199
198
200
199
1. Export selected features as a feature retrieval specification.
201
200
202
-
A feature retrieval specification is a portable definition of the feature list that's associated with a model. It can help streamline ML model development and operationalization. It will become an input to the training pipeline that generates the training data. Then, it will be packaged with the model.
201
+
A feature retrieval specification is a portable definition of the feature list that's associated with a model. It can help streamline the development and operationalization of a machine learning model. It will become an input to the training pipeline that generates the training data. Then, it will be packaged with the model.
203
202
204
203
The inference phase uses the feature retrieval to look up the features. It becomes a glue that integrates all phases of the machine learning lifecycle. Changes to the training/inference pipeline can stay at a minimum as you experiment and deploy.
Copy file name to clipboardExpand all lines: articles/machine-learning/tutorial-get-started-with-feature-store.md
+3-4Lines changed: 3 additions & 4 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -16,9 +16,9 @@ ms.custom: sdkv2, build-2023
16
16
17
17
# Tutorial 1: Develop and register a feature set with managed feature store (preview)
18
18
19
-
This tutorial series shows how features seamlessly integrate all phases of the machine learning (ML) lifecycle: prototyping, training, and operationalization.
19
+
This tutorial series shows how features seamlessly integrate all phases of the machine learning lifecycle: prototyping, training, and operationalization.
20
20
21
-
You can use Azure Machine Learning managed feature store to discover, create, and operationalize features. The ML lifecycle includes a prototyping phase, where you experiment with various features. It also involves an operationalization phase, where models are deployed and inference steps look up feature data. Features serve as the connective tissue in the ML lifecycle. To learn more about basic concepts for managed feature store, see [What is managed feature store?](./concept-what-is-managed-feature-store.md) and [Understanding top-level entities in managed feature store](./concept-top-level-entities-in-managed-feature-store.md).
21
+
You can use Azure Machine Learning managed feature store to discover, create, and operationalize features. The machine learning lifecycle includes a prototyping phase, where you experiment with various features. It also involves an operationalization phase, where models are deployed and inference steps look up feature data. Features serve as the connective tissue in the machine learning lifecycle. To learn more about basic concepts for managed feature store, see [What is managed feature store?](./concept-what-is-managed-feature-store.md) and [Understanding top-level entities in managed feature store](./concept-top-level-entities-in-managed-feature-store.md).
22
22
23
23
This tutorial is the first part of a four-part series. Here, you learn how to:
24
24
@@ -34,8 +34,7 @@ This tutorial series has two tracks:
34
34
* The SDK-only track uses only Python SDKs. Choose this track for pure, Python-based development and deployment.
35
35
* The SDK and CLI track uses the Python SDK for feature set development and testing only, and it uses the CLI for CRUD operations (create, update, and delete). This track is useful in continuous integration and continuous delivery (CI/CD) or GitOps scenarios, where CLI/YAML is preferred.
36
36
37
-
> [!IMPORTANT]
38
-
> This feature is currently in public preview. This preview version is provided without a service-level agreement, and we don't recommend it for production workloads. Certain features might not be supported or might have constrained capabilities. For more information, see [Supplemental Terms of Use for Microsoft Azure Previews](https://azure.microsoft.com/support/legal/preview-supplemental-terms/).
0 commit comments