# facebookresearch/StarSpace

Learning embeddings for classification, retrieval and ranking.

Repository: https://github.com/facebookresearch/StarSpace
Canonical: https://ross.abutalabs.com/products/starspace
Language: C++
License: MIT
License Family: permissive
Archived: true
Last push: 2022-12-04T04:02:21+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3353, "days_push": 1368, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3952, forks 524 (observed 2026-08-28T04:08:30.439329+00:00)

## What it is
StarSpace is a general-purpose neural model from Facebook Research that learns entity embeddings for classification, retrieval, ranking, and similarity tasks. It maps objects of different types into a common vector space so they can be compared and ranked against each other.

## Use cases
- learn word, sentence, or document embeddings
- rank documents or entities for information retrieval
- classify text or other labeled data
- learn sentence or document similarity metrics
- build content-based or collaborative filtering recommendations
- embed multi-relational knowledge graphs like Freebase
- classify or retrieve images using ResNet features

## When to choose
- you need a fast, general-purpose embedding model for ranking or classification across mixed entity types
- you want a lightweight C++ alternative to deep learning frameworks for embedding learning
- you need to compare entities of different types in a shared vector space

## When to avoid
- you need state-of-the-art transformer-based embeddings for NLP tasks
- you want an actively developed project with frequent updates
- you need GPU-accelerated training or deep integration with modern ML ecosystems

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, nlp, search-engine
- domain: machine-learning
- platform: windows, cpp, python
- tags: embeddings, entity-embeddings, ranking, recommendation, text-classification, similarity-learning, facebook-research, natural-language-processing, search, linux, macos

## Member repositories
- facebookresearch/StarSpace (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:30.439329+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 2026-08-29T18:24:31.329528+00:00, confidence not recorded.
  - readme: https://github.com/facebookresearch/StarSpace (fetched 2026-08-28T04:08:30.439329+00:00, sha dda4109585f4)
- Data as of 2026-08-30T08:39:29.467469+00:00.