We will now distribute our two document files between two different sets:
-
Documents in the training set are used to train the model. In other words, annotating these documents affects the way the model behaves.
-
Documents in the validation set are used by the model to assess its own performance. The model considers these documents "gold data," or, data it knows to be true because "we said so." Throughout the project, the model compares its predictions for each document in the validation set with that document's actual annotations. It then uses this comparison to calculate precision, recall, and F1 measure scores. For more information on these metrics, see the Adaptation Studio User Guide.
-
Select Add Documents from the project navigation bar.
-
Drag or browse for the document files in the Training Documents folder in the tutorial file package. Wait for the files to upload.
-
Select Add for training.
-
Drag or browse for the document files in the Validation Documents folder in the tutorial file package. Wait for the files to upload.
-
Select Add for validation.