| Concept | Definition |
|---|---|
| Project | A project in Generative AI Lab resembles a set of tasks that need to be annotated and/or reviewed by users in order to extract structured data and/or to train a DL model.Think of it as a factory assembly line for producing labels. For jumpstarting annotations on a project preannotations generated by existing models and/or rules can be used. |
| Project configuration | Specifies the type of documents that will be annotated as well as the labels, classes and relations which will be used for annotation. A project configuration can also reuse existing labels, classes and relations from pre-trained models or rules. |
| Model | In the context of the Annotation Lab use, the term model refers to a DL model build using John Snow Labs NLP libraries. |
| Predictions | Annotations automatically generated by Spark NLP models or user defined Spark NLP rules. |
| Completions | A series of annotations manually created or copied from automatic predictions and edited/validated by human annotators. |
| Task | A document that needs to be annotated by an annotator with or without the use of preannotation. |
PREVIOUSRelease Notes