- Release Notes
February 2023
Released in Endpoints + DocumentUnderstanding + Data Extraction ML Packages | v23.2.0
What's New & Improvements
A new version of the Out-of-the-box Pre-trained ML Packages (23.1.0) and their public endpoints has been released, now using cutting edge LayoutLM Transformers based architecture, which is more powerful and increases accuracy overall, especially on column fields (tables).
This improvement has made the Out-of-the-box Pre-trained ML Packages more powerful, meaning that you may experience longer latency for training and for predictions.
For all situations where latency is critical (e.g.: attended scenarios) we recommend deploying the models as ML Skills using a GPU.
We have improved how the scores are calculated after Training/Evaluation/Full pipelines to provide a separate score for each column field. Before this improvement, F1 scores were calculated as a whole, for all column fields taken together.
An upcoming removal is announced for the Manual edits feature used in the model evaluation. More information here.
Known Issues
The project import from AI Center is currently disabled. We are actively working on this and expect to have it reenabled by the end of March.
Erratum 8 May 2023
Known Issue
Fatal Python error: Segmentation fault
is received when running a Full or Training Pipeline. We recommend using the ML Packages with v23.4 until this bug is fixed.
Erratum 20 April 2023
Overall score for all pipelines is now an Accuracy. Previously it was an F1 score. The evaluation artefacts in AI Center still contain both accuracy and F1 score, for backwards comparability.