First Steps¶
Install and verify AutoRAG-Research.
Install¶
pip install autorag-research
For development installation:
git clone https://github.com/NomaDamas/AutoRAG-Research.git
cd AutoRAG-Research
uv sync --all-extras
Requirements¶
- Python 3.10+
- PostgreSQL 18 client tools (
pg_restore-18for database restore)
Verify PostgreSQL tools:
pg_restore-18 --version
Start PostgreSQL¶
docker-compose up -d
Initialize Configuration¶
autorag-research init
Creates configs/ directory with templates.
Verify¶
autorag-research show ingestors
You should see available dataset ingestors:
| Name | Description |
|---|---|
| beir | BEIR benchmark datasets |
| mteb | MTEB retrieval tasks |
| ragbench | RAGBench benchmark |
| vidorev3 | ViDoRe V3 visual documents |
Next¶
- Text Retrieval - Run your first benchmark
- Concepts - Understand the system