# VowpalWabbit/vowpal_wabbit

Vowpal Wabbit is a machine learning system which pushes the frontier of machine learning with techniques such as online, hashing, allreduce, reductions, learning2search, active, and interactive learning.

Repository: https://github.com/VowpalWabbit/vowpal_wabbit
Canonical: https://ross.abutalabs.com/products/vowpal_wabbit
Homepage: https://vowpalwabbit.org
Language: C++
License: NOASSERTION
License Family: other
Topics: c-plus-plus, machine-learning, online-learning, contextual-bandits, reinforcement-learning, active-learning, learning-to-search, cpp
Last push: 2026-08-26T18:58:34+00:00

## Health v2 (maintenance only)
Score: 90/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 73, longevity 100
- inputs: {"age_days": 6242, "days_push": 7, "days_rel": 182, "gap_med": 1, "n_releases_24m": 2}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 8706, forks 1926 (observed 2026-08-28T04:10:24.323189+00:00)

## What it is
Vowpal Wabbit is a fast, open-source machine learning system focused on online, active, and interactive learning, including contextual bandit reinforcement learning and learning-to-search. It offers a C++ core with CLI, Python, C#, and Java bindings, and scales via hashing and bounded-memory streaming training.

## Use cases
- train models on streaming data that doesn't fit in memory
- build a contextual bandit recommender or personalization system
- run online supervised learning from the command line
- do active learning to minimize labeling costs
- solve joint prediction problems with learning to search
- extreme multi-class classification with many labels
- evaluate policies offline before deploying a bandit

## When to choose
- you need fast online/incremental learning with a small memory footprint
- you're building contextual bandit or interactive RL systems
- you want battle-tested ML research algorithms with CLI and Python bindings

## When to avoid
- you need deep learning with GPUs or neural architectures
- you want a high-level scikit-learn-style API with extensive preprocessing
- your team can't work with a custom input format and CLI-centric workflow

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, reinforcement-learning, cli
- domain: machine-learning, reinforcement-learning, developer-tools
- platform: windows, python, cpp, cli, cross-platform
- tags: online-learning, contextual-bandits, active-learning, learning-to-search, hashing-trick, extreme-classification, supervised-learning, algorithms, linux, macos

## Member repositories
- VowpalWabbit/vowpal_wabbit (main) score 90

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:24.323189+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 2026-08-29T17:25:22.804525+00:00, confidence not recorded.
  - readme: https://github.com/VowpalWabbit/vowpal_wabbit (fetched 2026-08-28T04:10:24.323189+00:00, sha 0e6449a9169d)
  - homepage: https://vowpalwabbit.org (fetched 2026-08-29T08:25:48.255289+00:00, sha a63c66597814)
  - site_page: https://vowpalwabbit.org/features.html (fetched 2026-08-29T08:25:48.264907+00:00, sha 6abe7f205320)
  - site_page: https://vowpalwabbit.org/docs/vowpal_wabbit/python/latest/index.html (fetched 2026-08-29T08:25:48.267017+00:00, sha 9b4f85809417)
- Data as of 2026-08-30T08:39:29.467469+00:00.