AmenRa/retriv

A Python Search Engine for Humans 🥸

57
/ 100
Established

Supports sparse (BM25/TF-IDF), dense (semantic), and hybrid retrieval modes, leveraging Numba for vectorized operations, Transformers for embedding generation, and Faiss for approximate nearest-neighbor search. Includes automatic tuning via Optuna for BM25 parameters and Faiss configuration, plus multi-language text preprocessing with configurable stemmers and tokenizers. Provides unified search interface across all retrieval types with batch processing capabilities for large-scale evaluation without memory constraints.

248 stars and 1,049 monthly downloads. Available on PyPI.

Maintenance 6 / 25
Adoption 17 / 25
Maturity 18 / 25
Community 16 / 25

How are scores calculated?

Stars

248

Forks

31

Language

Python

License

MIT

Last pushed

Dec 18, 2025

Monthly downloads

1,049

Commits (30d)

0

Dependencies

19

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/embeddings/AmenRa/retriv"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.