# databricks/lilac

Curate better data for LLMs

Repository: https://github.com/databricks/lilac
Canonical: https://ross.abutalabs.com/products/lilac
Homepage: http://lilacml.com
Language: Python
License: Apache-2.0
License Family: permissive
Topics: artificial-intelligence, data-analysis, dataset-analysis, unstructured-data
Archived: true
Last push: 2024-03-19T12:41:30+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 89
- inputs: {"age_days": 1259, "days_push": 897, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1072, forks 105 (observed 2026-08-28T04:03:28.686093+00:00)

## What it is
Lilac is an open-source tool for exploring, curating, and quality-controlling datasets used for training, fine-tuning, and monitoring LLMs. It provides both a web UI and Python API, running on-device with open-source LLMs for search, filtering, clustering, and annotation of unstructured data.

## Use cases
- curate fine-tuning datasets for LLMs
- find and remove PII from training data
- deduplicate large text datasets before training
- explore and search millions of documents with semantic search
- cluster and label unstructured text data with LLMs
- monitor dataset quality changes over time
- visualize pre-training data distributions

## When to choose
- you need to clean and curate large unstructured text datasets for LLM training or fine-tuning
- you want an interactive UI plus Python API for dataset exploration
- you need on-device processing with open-source LLMs for data quality tasks like PII removal and deduplication

## When to avoid
- you need a fully managed enterprise data platform with support guarantees
- your data is structured tabular data rather than unstructured text
- you require active development and frequent updates, as the latest release is from March 2024

## Facets
- artifact type: library
- maturity: maintenance
- function: data-science, etl, search-engine, nlp, machine-learning, data-visualization
- domain: artificial-intelligence, large-language-models, data-science, machine-learning
- platform: python, self-hosted, cross-platform
- tags: dataset-curation, llm-data-quality, data-exploration, fine-tuning-data, pii-detection, semantic-search, clustering, web-ui, data-engineering, web-server

## Member repositories
- databricks/lilac (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:28.686093+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:53:41.075383+00:00, confidence not recorded.
  - readme: https://github.com/databricks/lilac (fetched 2026-08-28T04:03:28.686093+00:00, sha 9f7261a876fa)
- Data as of 2026-08-30T08:39:29.467469+00:00.