Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

OpenAutoNLU: Open Source AutoML Library for NLU

Grigory Arshinov, Aleksandr Boriskin, Sergey Senichev, Ayaz Zaripov, Daria Galimzianova, Daniil Karpov, Leonid Sanochkin

TL;DR

OpenAutoNLU is an open-source automated machine learning library for natural language understanding (NLU) tasks, covering both text classification and named entity recognition, that introduces data-aware training regime selection that requires no manual configuration from the user.

Abstract

OpenAutoNLU is an open-source automated machine learning library for natural language understanding (NLU) tasks, covering both text classification and named entity recognition (NER). Unlike existing solutions, we introduce data-aware training regime selection that requires no manual configuration from the user. The library also provides integrated data quality diagnostics, configurable out-of-distribution (OOD) detection, and large language model (LLM) features, all within a minimal lowcode API. The demo app is accessible here https://openautonlu.dev.

OpenAutoNLU: Open Source AutoML Library for NLU

TL;DR

OpenAutoNLU is an open-source automated machine learning library for natural language understanding (NLU) tasks, covering both text classification and named entity recognition, that introduces data-aware training regime selection that requires no manual configuration from the user.

Abstract

OpenAutoNLU is an open-source automated machine learning library for natural language understanding (NLU) tasks, covering both text classification and named entity recognition (NER). Unlike existing solutions, we introduce data-aware training regime selection that requires no manual configuration from the user. The library also provides integrated data quality diagnostics, configurable out-of-distribution (OOD) detection, and large language model (LLM) features, all within a minimal lowcode API. The demo app is accessible here https://openautonlu.dev.
Paper Structure (25 sections, 2 figures, 6 tables)

This paper contains 25 sections, 2 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Background
  3. OpenAutoNLU
  4. Structure
  5. Data quality
  6. Named Entity Recognition.
  7. Optimized-inference-ready model serialization
  8. Text Classification Training Pipeline
  9. Data Quality
  10. Objective and search strategy.
  11. Data-level optimisation
  12. OOD detection optimisation
  13. LLM-based test set generation
  14. Experiments
  15. OOD Unaware Regime
  16. ...and 10 more sections

Figures (2)

  • Figure 1: Text Classification Training Pipeline and Inference Pipeline flow
  • Figure 2: This graph illustrates the ratio between performance in macro F1 on classification tasks and time in seconds took to train the solution. Measures are averaged between four different text classification datasets.