Documentation
HomeBlogsGet Started
  • ๐Ÿ’กIntroduction
  • ๐Ÿš€Unitlab Annotate
  • Workspace Management
    • ๐Ÿ Create a Workspace
    • ๐Ÿ‘จโ€๐Ÿ‘ฉโ€๐Ÿ‘งโ€๐Ÿ‘ฆInvite Members
    • โ›ณRole-Based Access
  • Project Management
    • ๐Ÿ“ŒSetup a Project
    • ๐Ÿ–Œ๏ธAnnotation
    • ๐Ÿ“‰Performance Analytics
  • Auto Labeling
    • ๐Ÿช„Segment Anything (SAM)
    • ๐Ÿค–Batch Auto-Annotation
    • โน๏ธCrop Auto-Annotation
  • Ai models
    • ๐Ÿง Model Integration
    • ๐Ÿ”งModel Management
    • ๐ŸŽกHow To Use?
  • AUTOMATION WORKFLOW
    • โ›“๏ธAutomation Workflow
  • Dataset Management
    • ๐Ÿ›ฐ๏ธRelease Datasets
    • โš™๏ธManage Datasets
    • ๐ŸŒ€Clone Datasets
  • CLI/Python SDK
    • ๐Ÿ› ๏ธGet started
    • ๐Ÿ’ปUnitlab CLI
    • ๐Ÿ›ก๏ธUnitlab Python SDK
  • How to integrate
    • ๐Ÿ”‘Create an API key
Powered by GitBook
On this page
  • Cloning a Dataset
  • Restoring a Project

Was this helpful?

  1. Dataset Management

Clone Datasets

Clone Datasets and create your own version

PreviousManage DatasetsNextGet started

Last updated 1 year ago

Was this helpful?

In Unitlab Annotate, datasets are released from projects, while projects can also be created from the released datasetsโ€”a process known as the cyclical workflow of Unitlab Annotate.

The cyclical workflow of Unitlab Annotate enables you to perform projects to datasets, datasets to projects, and so on, engaging you until your datasets reach a high-quality level. This workflow contains actions such as "," "," and "."

If you missed learning how to release and manage datasets, you can learn about them on the "" and "" pages, respectively.

In Unitlab Annotate, cloning datasets is a part of the cyclical workflow and is used for two main purposes:

  • : Clone other's public datasets and make your own version.

  • : Restore a project with specific version of the dataset.

Cloning a Dataset

When you click the "Clone Dataset" button for the dataset you want to clone, it asks whether to clone with annotations or without annotations as follows:

If you prefer to clone a dataset with annotations, then both the annotations and data remain in the project creation. Otherwise, only the data (images) without annotations is available in the project creation.

When you click "confirm," a project creation process starts as shown in the screenshot below:

Restoring a Project

Each new dataset release from your project creates a new version of the dataset. You can restore the project using any specified version of the dataset.

To restore a project, go to the dataset you want to restore as a project and select the preferred version and click "Clone Dataset" located in the upper right corner of the dataset page.

The project creation steps are the same as those we learned on the "" page. Learn how to create projects from scratch on this page.

When you select a dataset version and click "Clone Dataset," the project creation begins with the selected version of annotations. The process for creating a project is standard and follows the same steps as those we learned in "."

๐ŸŒ€
Setup a Project
๐Ÿ“ŒSetup a Project
Setup a Project
Release
Manage
Release Datasets
Manage Datasets
๐Ÿ›ฐ๏ธRelease Datasets
โš™๏ธManage Datasets
Clone Datasets
Cloning a Dataset
Restoring a Project
Unitlab Annotate: Dataset Page - "Clone Dataset" Button
Unitlab Annotate: Dataset page - Cloning a Dataset
Unitlab Annotate: Dataset Page - "Cloning a Dataset with or without Annotations" Pop-Up
Unitlab Annotate: A Cloned Dataset - Project Creation
Unitlab Annotate: Clone Dataset with a Version