Forum Discussion
Post ProdigyMigrating Azure Machine Learning pipelines to Fabric
Hi,
in order to move or convert or re-create some pipelines of Azure ML in Fabric,
which is the corresponding Fabric item respect to a ML pipeline?
Thanks
Hi pmscorca ,
In Microsoft Fabric, the closest equivalent to Azure ML pipelines would be Data Pipelines (under the Data Factory experience) and Notebooks for more custom ML workflows. While Fabric doesn’t yet have a full ML pipeline orchestration layer like Azure ML, you can re-create many steps using:
- Data Pipelines for orchestrating data movement and transformation
- Notebooks (PySpark or Python) for training, scoring, and evaluating models
- Lakehouse as your central data storage
- MLflow integration (coming or partially available) for experiment tracking
If your AML pipelines include model training and deployment, you might need to combine Fabric with Azure ML or other services for now, depending on your use case.
If my response resolved your query, kindly mark it as the Accepted Solution to assist others. Additionally, I would be grateful for a 'Kudos' if you found my response helpful.
1 Reply
- burakkaragoz
Super User
Hi pmscorca ,
In Microsoft Fabric, the closest equivalent to Azure ML pipelines would be Data Pipelines (under the Data Factory experience) and Notebooks for more custom ML workflows. While Fabric doesn’t yet have a full ML pipeline orchestration layer like Azure ML, you can re-create many steps using:
- Data Pipelines for orchestrating data movement and transformation
- Notebooks (PySpark or Python) for training, scoring, and evaluating models
- Lakehouse as your central data storage
- MLflow integration (coming or partially available) for experiment tracking
If your AML pipelines include model training and deployment, you might need to combine Fabric with Azure ML or other services for now, depending on your use case.
If my response resolved your query, kindly mark it as the Accepted Solution to assist others. Additionally, I would be grateful for a 'Kudos' if you found my response helpful.