llm-annotator 0.4.0

An easy-to-extend LLM annotator for robust, resumable data annotation.

A simple, extensible LLM Annotator

This repository provides a small, resumable framework for annotating datasets with LLMs (via vllm). Below is a minimal usage example showing how to instantiate the Annotator class and run a short annotation job.

Installation

Recommended:

uv add llm-annotator

or

pip install llm-annotator

Installing flash-infer for your version (eg CUDA12.8)

uv pip install flashinfer-python flashinfer-cubin
# JIT cache package (replace cu129 with your CUDA version: cu128, cu129, or cu130)
uv pip install flashinfer-jit-cache --index-url https://flashinfer.ai/whl/cu128

Usage

See examples/ for usage examples.

Testing

pytest -q