Custom AI

Datasets Page

Content is machine translated from English by Phrase Language AI.

Created datasets are presented in a table on the Datasets page. The table can be ordered by clicking on column headings. Clicking on the dataset name opens the Overview for the selection. Hover over the right of a dataset name to rename edit.svg that dataset.

The More menu 3dots_icon.jpeg can also be used to open the Overview, reusing the dataset and filters for a new dataset or deleting the dataset.

Dataset Overview

The overview provides information about the cleaning with tabs for reviewing the applied filters and a history of the cleaning of the selected dataset. The distribution of cleaning filters is displayed along with all metadata.


  • Overview

    Displays the overview of that dataset and the distribution of cleaning filters.

  • Cleaning filters

    Displays the results of the filters selected for the initial training.

    To change the filters, follow these steps:

    1. Click Edit filters .

      The Edit cleaning filters page opens.

    2. Click Edit on a filter that requires a change.

      The filter options open.

    3. Make required changes and click Save.

      The changes are displayed and the Reclean button is activated.

    4. Click Reclean to update the cleaning.

  • Cleaning history

    A table of cleanings is presented.

Evaluation dataset

The number of segments selected for custom MT model evaluation is determined by the lesser of the following two criteria:

  • 10% of the total dataset.

  • A fixed count of 3,000 segments.

These segments are exclusively for evaluation and are not used during the training phase of the model.

Was this article helpful?

Sorry about that! In what way was it not helpful?

The article didn’t address my problem.
I couldn’t understand the article.
The feature doesn’t do what I need.
Other reason.

Note that feedback is provided anonymously so we aren't able to reply to questions.
If you'd like to ask a question, submit a request to our Support team.
Thank you for your feedback.