PyPI page
Home page
Author:
None
License:
MIT
Summary:
ALIRA classifies a large text corpus according to a natural-language query when exhaustive LLM evaluation is impractical. It iteratively discovers relevant documents using active learning, LLM validation, and classifier refinement.
Latest version:
0.1.0
Required dependencies:
numpy
|
openai
|
pandas
|
pydantic
|
scikit-learn
Optional dependencies:
build
|
epfl-data-index
|
pyarrow
|
pytest
|
python-dotenv
Downloads last day:
2
Downloads last week:
13
Downloads last month:
164