Pipelines
The Pipelines menu is used to define a series of steps required to build, train, evaluate the created model. On the Portal Cloudeka Deka MLOps Service menu page, select the Pipelines menu.
The Pipelines menu has three main features:
Upload Pipeline This feature allows you to upload a new pipeline definition to the Portal Cloudeka Deka MLOps Service environment. A pipeline definition is typically a YAML or JSON file that describes the workflow, including the sequence of steps, components, and parameters required for execution. By using the "Upload Pipeline" option, you can introduce new workflows to automate their machine learning processes. Once uploaded, the pipeline can be accessed, edited, and executed as needed.
Refresh The "Refresh" option is utilized to update the list of pipelines displayed in the menu. This is particularly useful when changes have been made to the pipelines, such as new uploads or modifications to existing pipelines. By clicking "Refresh," you ensure that the information you see is current, allowing you to easily access the latest versions of pipelines available in the Portal Cloudeka Deka MLOps Service environment.
Delete This feature enables you to permanently remove a selected pipeline from the system. Deleting a pipeline is useful when it is no longer needed or if it has become obsolete. You can select the pipeline you wish to delete and confirm the action, which will remove it from the list of available pipelines. It’s important to note that once deleted, the pipeline cannot be recovered without re-uploading or recreating it.
In the Pipelines menu of the Portal Cloudeka Deka MLOps Service, there are two types of pipelines: Private and Shared. Each type serves a different purpose and is designed to accommodate various collaboration and access needs within a team or organization.
a. Private Private pipelines are accessible only to the user who created them. This type of pipeline is ideal for individual experimentation and development, allowing you to work on your workflows without interference from others. You can test and refine your pipelines in a secure environment, ensuring that sensitive data or proprietary algorithms remain confidential. Once you or a user is satisfied with your/their private pipeline, you/they can choose to share it with others if desired.
b. Shared Shared pipelines, on the other hand, are accessible to multiple users within the Kubeflow environment. This type of pipeline is beneficial for collaborative projects where team members need to work together on the same workflows. By sharing pipelines, users can leverage each other's expertise, contribute to the development process, and ensure consistency across different stages of a project. Shared pipelines facilitate better communication and collaboration, making it easier to manage complex machine learning workflows that involve multiple stakeholders.
Last updated