README.md

hirundo ·

The Hirundo Python SDK lets you:

• Launch and monitor LLM behavior unlearning runs.
• Run LLM behavior evaluations for bias, hallucination, and prompt injection.
• Run dataset QA for ML datasets (classification, object detection, and more).
• Fetch QA results as pandas or polars DataFrames.

This SDK requires access to a Hirundo server (SaaS, VPC, or on-prem).

Requirements

• Python 3.10, 3.11, 3.12, or 3.13 (CPython).
• A Hirundo API key.

Installation

pip install hirundo

Optional extras:

• LLM behavior unlearning (Transformers + PEFT): pip install hirundo[transformers]
• Dataset QA or LLM behavior eval results as DataFrames: pip install hirundo[pandas] or pip install hirundo[polars]

If you want to install from source, clone this repository and run:

pip install .

Configure API access

You can set environment variables directly or use the CLI helper:

hirundo setup

This writes API_KEY (and optionally API_HOST) to .env in the current directory or ~/.hirundo.conf.

Quickstart examples

The full quickstart examples now live in the Sphinx docs so they can be linted, formatted, and type-checked as real Python files. See the examples embedded in docs/index.rst, which are sourced from docs/*.py files.

Supported dataset storage

• Amazon S3
• Google Cloud Storage (GCS)
• Git repositories with LFS (GitHub, Hugging Face)

Further documentation

• Documentation site: https://docs.hirundo.io/
• Example notebooks: notebooks/