Edit a learning instance
Edit a learning instance to change its name, description, add additional training documents, or include additional fields for extraction.
You can edit learning instances that are in the staging environment. To edit instances in the production environment, first set the learning instance to the staging environment.
- On the Learning Instances page, click a learning instance.
- Click the Edit icon to edit the learning instance.
- Optional: Update the description field.
Change or update the learning instance
When you change the name of a learning instance, the following changes occur:
- A separate folder with the new learning instance name is created in the
configured output directory to store the .csv files of processed
The previous folder containing the .csv files before the learning instance name was changed remains as is in the output folder.
- When you migrate a learning instance with an updated name using overwrite or remove options, it is updated in the source environment after the .iqba file is imported.
- An entry in the audit logs is created for this activity.
- A separate folder with the new learning instance name is created in the configured output directory to store the .csv files of processed documents.
- Optional: Click Browse to select additional training documents.
- Optional: Select additional fields and table columns to add to the learning instance for data extraction.
Use the New groups creation setting field to restrict
creation of new groups or set the threshold for the number of groups.
Note: The New groups creation setting option is not applicable for IQ Bot [Local Device] and IQ Bot Extraction packages.
Use the Default validations group option to select a
default group for document validations.
All new documents use the custom logic and validations defined in the default group. Validation rules from this default group are applied when a document does not belong to any of the existing group.
- Click Save to view a confirmation message.
Click Yes, proceed with field addition to upload,
analyze, and classify the selected training documents (if any) that were added
to the learning instance.
The classification of documents remain unaffected by the newly added fields.
All additional fields become available in all the bots and appear as optional fields in the selected learning instance. The bots must be retrained for these fields.