Describe 

Here is shown a brief description of the model. We introduce the different fields in the following: 

• Status: status of the model. Can be: 

  • NEW: after the model is created. If nothing went wrong, this status is quickly updated to READY. 

  • READY: this means the model is ready to be used for any of the available actions. 

  • TRAINING: the model is still training and needs to finish the process for updating its status to READY. 

• Language: the language defined during model creation. 

• Description: a brief description of the model and used as its name identifier within the app. 

• Labels: the labels and their label texts associated defined during model creation. 

• Dataset: the dataset associated with the model. 

• Xval_macro_f1: if the model has already been trained with some labeled instances this will show the macro average F1-score on that data. 

• Model Type: the model type selected during model creation. 

• Is multi-label: if the model has been defined as multi-label. Otherwise, will be multi-class (or binary if only 2 classes were defined for the task). 

• Embedding model: the underlying model selected during model creation. Can be Symanto Fast or Symanto. 

Download labeled examples 

Request to download a file with all the text instances that the user had annotated until that moment. 

Download unlabeled examples 

Request to download a file of the dataset associated to the model annotated with the model’s prediction. Each text instance of the unlabeled dataset will be annotated with the most probable label and the label probability distribution inferred by the model. 

Delete 

Remove the model from the DB. This action will be irreversible.