# microsoft/Semi-supervised-learning

A Unified Semi-Supervised Learning Codebase (NeurIPS'22)

Repository: https://github.com/microsoft/Semi-supervised-learning
Canonical: https://ross.abutalabs.com/products/semi-supervised-learning
Homepage: https://usb.readthedocs.io
Language: Python
License: MIT
License Family: permissive
Topics: classification, semi-supervised-learning, transformer, computer-vision, deep-learning, machine-learning, natural-language-processing, pytorch, audio-classification, low-resource, semisupervised-learning
Last push: 2026-07-26T09:08:25+00:00

## Health v2 (maintenance only)
Score: 65/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 94, release rhythm 8, longevity 100
- inputs: {"age_days": 1581, "days_push": 38, "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 1588, forks 220 (observed 2026-08-28T04:05:07.987508+00:00)

## What it is
USB (Unified Semi-supervised learning Benchmark) is a PyTorch-based codebase from Microsoft for semi-supervised learning across computer vision, NLP, and audio classification. It provides implementations of popular SSL algorithms with a unified framework and benchmark results.

## Use cases
- train semi-supervised image classifiers with limited labels
- apply SSL algorithms to text classification
- run audio classification with few labeled examples
- benchmark FixMatch, FlexMatch, and other SSL methods
- experiment with transformers in low-resource settings

## When to choose
- you need a unified framework to compare SSL algorithms
- you want reproducible benchmarks across CV, NLP, and audio
- you work in PyTorch and need ready-made SSL implementations

## When to avoid
- you need fully supervised training pipelines only
- you require production deployment tooling rather than research code
- you work outside PyTorch

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning
- domain: machine-learning, computer-vision
- platform: python
- tags: semi-supervised-learning, pytorch, classification, benchmark, low-resource, natural-language-processing, audio

## Member repositories
- microsoft/Semi-supervised-learning (main) score 65

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:07.987508+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 2026-08-30T03:55:04.307904+00:00, confidence not recorded.
  - readme: https://github.com/microsoft/Semi-supervised-learning (fetched 2026-08-28T04:05:07.987508+00:00, sha e6ccf6feb2a1)
- Data as of 2026-08-30T08:39:29.467469+00:00.