Creating and managing a project

A Project is a set of data (documents or pages with various annotations used for training and evaluation) and models (a system that you train on the training data to make the predictions on other documents or pages not seen before).

Create a new Project

1. Start a new project

   Click the Create a new Project panel on the H2O Document AI - Publisher homepage

   

2. Name the project

   Enter the name of your project

3. Set auto-purge

   Toggle whether to auto-purge your training artifacts

4. Set scheduled deletion

   Toggle whether to schedule the deletion of your project along with its resources

5. Add a description

   (Optional) Add a description for your project

6. Add collaborators

   Input any collaborators you want to work with on this project (if none, then leave empty)

7. Upload your documents

   Upload the desired documents, images, or compressed files

8. Create the project

   Click Create project

You will find your newly created project under Projects on the left navigation bar. You will also see the artifact uploaded under the Documents sets and Annotation sets sections on your Project home page.

info

Depending upon the size of the uploaded file, it might take some time to populate the Documents sets and Annotation sets sections with the number of documents and the number of pages. You can check the status of the upload by looking at the Status column on the Document sets, Annotation sets, or Jobs pages. A job that is running is pending or processing. A finished job is done or ready.

You can also check the status of a project by opening the background processes tab. You access this by clicking the bell icon next to your account name.

Auto-purge

Automatically purge your training artifacts. Set this option when creating or editing your project. This feature is enabled by your system administrator. The retention period is set by your system administrator, too.

Scheduled deletion

Schedule the deletion of your whole project and all of its resources. By toggling on this feature while creating or editing your project, you can set the date you want to erase your project.

Project collaboration

You can add other users (collaborators) to work on your project by providing their email addresses in the Edit Project panel in the upper navigation bar.

The project creator is the owner, and ownership cannot be transferred. Only the owner can add or remove collaborators and open the Edit Project panel. Aside from these owner-specific privileges, collaborators have the same permissions as the owner, including the ability to work with document and annotation sets and publish pipelines.

The UI manages collaborators using a picker component, and role bindings are made exclusively through email. The UI verifies that the email exists within your organization, but it does not provide autocomplete. If the email is invalid, an error response will be displayed.

Project panels

The project panels show overviews of each page. You can access each page by clicking See All on this panel or by accessing it via the left navigation bar.

Document sets

See the most recent updates and scroll through older ones. Select "See All" to access the Document sets page.

Annotation sets

See the most recent updates and scroll through older ones. Select "See All" to access the Annotation sets page.

Models

See the most recent updates and scroll through older ones. Select "See All" to access the Models page.

Jobs

See the most recent updates and scroll through older ones. Select "See All" to access the Jobs page.

Project jobs

Interact with your project from the upper navigation bar.

Delete project

Delete your project. You will be prompted to acknowledge that the act of deletion is destructive and irreversible.

Export Project images

Export all or selected images from your project. Exporting lets you work with annotation sets offline.

1. Select the images to export

   Choose which annotation set images to export by clicking the box next to their names

2. Export the images

   Click Export

When the job is finished running, the images will be downloaded to your local computer.

Import Post Processor

Import a custom Post Processor.

1. Name the post processor

   Provide a name for your post processor

2. Add a description

   (Optional) Provide a description of your post processor

3. Add the zipped directory

   Drag and drop the zipped post processor directory or browse your local files

4. Import the post processor

   Click Import

Import Model

Import a previously built model that you have saved to your local computer.

1. Name the model

   Provide a name for your imported model

2. Add a description

   (Optional) Provide a description for your model

3. Add the zipped model file

   Drag and drop the file or browse your local files for the zipped model file

4. Import the model

   Click Import

Import Annotation set

Import a JSON file of a previously annotated annotation set that you have saved to your local computer.

1. Name the annotation set

   Provide a name for the imported annotation set

2. Add a description

   (Optional) Provide a description for your annotation set

3. Add the JSON file

   Drag and drop the JSON file or browse your local files for the file you want to import

4. Import the annotation set

   Click Import

Import Document set

This will import a set of documents.

1. Name the document set

   Provide a name for the document set

2. Add a description

   (Optional) Provide a description for your document set

3. Copy attributes

   Select if you want to copy the attributes from an available attribute set

4. Upload your documents

   Upload the desired documents, images, or compressed files. You can either drag and drop the zip file or browse for it

Once the document is imported, it will appear in the Document sets page and an entry will appear in the Annotation sets page.

---

Feedback

• Submit and view feedback for this page
• Send feedback about H2O Document AI to cloud-feedback@h2o.ai