# snorkel-team/snorkel

A system for quickly generating training data with weak supervision

Repository: https://github.com/snorkel-team/snorkel
Canonical: https://ross.abutalabs.com/products/snorkel
Homepage: https://snorkel.org
Language: Python
License: Apache-2.0
License Family: permissive
Topics: machine-learning, ai, weak-supervision, labeling, data-science, python, snorkel, training-data, data-augmentation, data-slicing
Last push: 2026-06-08T19:59:20+00:00

## Health v2 (maintenance only)
Score: 62/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 86, release rhythm 8, longevity 100
- inputs: {"age_days": 3841, "days_push": 86, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 6001, forks 858 (observed 2026-08-28T04:09:34.163648+00:00)

## What it is
Snorkel is a Python library for programmatically building and managing training data using weak supervision, letting users write labeling functions instead of hand-labeling data. It also supports data augmentation and data slicing, and originated as a Stanford research project.

## Use cases
- generate training data without manual labeling
- write labeling functions for weak supervision
- programmatically label a large dataset
- augment training data for machine learning
- create data slices to evaluate model performance
- denoise noisy labels from multiple labeling sources

## When to choose
- you need to label large datasets quickly without hand-annotation
- you have domain heuristics that can be encoded as labeling functions
- you want research-backed weak supervision modeling in Python

## When to avoid
- you need an end-to-end commercial platform with a UI (consider Snorkel Flow)
- you expect active feature development, as the team has shifted focus to Snorkel Flow
- you need fully automated labeling with no heuristic input

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, data-science, etl
- domain: machine-learning, data-science, artificial-intelligence
- platform: python, cross-platform
- tags: weak-supervision, data-labeling, training-data, data-augmentation, data-slicing, labeling-functions

## Member repositories
- snorkel-team/snorkel (main) score 62

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:34.163648+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 2026-08-29T17:49:48.984961+00:00, confidence not recorded.
  - readme: https://github.com/snorkel-team/snorkel (fetched 2026-08-28T04:09:34.163648+00:00, sha 39c9a0a49113)
  - registry_pypi: https://pypi.org/pypi/snorkel/json (fetched 2026-08-29T08:45:47.137898+00:00, sha df57fbe0a8c4)
- Data as of 2026-08-30T08:39:29.467469+00:00.