# facebookresearch/SentEval

A python tool for evaluating the quality of sentence embeddings.

Repository: https://github.com/facebookresearch/SentEval
Canonical: https://ross.abutalabs.com/products/senteval
Language: Python
License: NOASSERTION
License Family: other
Archived: true
Last push: 2024-03-19T16:54:22+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": 3394, "days_push": 897, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2110, forks 307 (observed 2026-08-28T04:06:14.205325+00:00)

## What it is
SentEval is a Python library from Facebook Research for evaluating the quality of sentence embeddings via a suite of downstream transfer tasks and linguistic probing tasks. It helps researchers benchmark general-purpose fixed-size sentence representations.

## Use cases
- evaluate quality of my sentence embeddings
- benchmark sentence encoders on downstream NLP tasks
- test what linguistic properties my sentence representations encode
- compare sentence embeddings on semantic textual similarity
- run standard NLP evaluation benchmarks like SNLI and STS

## When to choose
- you need standardized benchmarks for sentence embedding quality
- you are researching general-purpose sentence representations
- you want probing tasks to analyze linguistic information in embeddings

## When to avoid
- you need to generate or train embeddings rather than evaluate them
- you need actively maintained tooling with recent NLP benchmarks
- you are not working in Python with PyTorch

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, nlp, benchmarking, testing
- domain: machine-learning
- platform: python
- tags: sentence-embeddings, evaluation-toolkit, transfer-tasks, probing-tasks, pytorch, research, natural-language-processing, linux, macos

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

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:14.205325+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 2026-08-30T02:53:44.690327+00:00, confidence not recorded.
  - readme: https://github.com/facebookresearch/SentEval (fetched 2026-08-28T04:06:14.205325+00:00, sha 508b6efed444)
- Data as of 2026-08-30T08:39:29.467469+00:00.