# TorchSSL/TorchSSL

A PyTorch-based library for semi-supervised learning (NeurIPS'21)

Repository: https://github.com/TorchSSL/TorchSSL
Canonical: https://ross.abutalabs.com/products/torchssl
Language: Python
License: MIT
License Family: permissive
Topics: semi-supervised-learning, toolkit, codebase, pytorch, machine-learning, computer-vision, deep-learning, self-supervised-learning
Last push: 2023-08-28T06:32:01+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 1786, "days_push": 1101, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1372, forks 185 (observed 2026-08-28T04:04:32.358100+00:00)

## What it is
TorchSSL is a PyTorch-based toolbox for semi-supervised learning that implements popular SSL algorithms like FixMatch, FlexMatch, MixMatch, and MeanTeacher for fair benchmarking. It is the official implementation of FlexMatch (NeurIPS 2021) but is now deprecated in favor of Microsoft's USB.

## Use cases
- reproduce semi-supervised learning benchmarks on CIFAR and SVHN
- compare SSL algorithms like FixMatch and FlexMatch under one codebase
- implement curriculum pseudo labeling for pseudo-label and UDA methods
- study FlexMatch source code from the NeurIPS 2021 paper
- train semi-supervised image classification models with limited labels

## When to choose
- you specifically need the original FlexMatch or TorchSSL reference implementation
- you want to replicate the NeurIPS 2021 benchmark results exactly

## When to avoid
- you are starting a new SSL project - use the successor USB instead
- you need maintained code with bug fixes or updated results
- you need SSL beyond computer vision, such as NLP or audio

## Facets
- artifact type: library
- maturity: abandoned
- function: machine-learning, deep-learning
- domain: machine-learning, computer-vision, deep-learning
- platform: python
- tags: semi-supervised-learning, pytorch, self-supervised-learning, benchmark, research-toolkit, deprecated, gpu

## Member repositories
- TorchSSL/TorchSSL (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:32.358100+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 2026-08-30T04:40:45.192289+00:00, confidence not recorded.
  - readme: https://github.com/TorchSSL/TorchSSL (fetched 2026-08-28T04:04:32.358100+00:00, sha cdb0755a0757)
- Data as of 2026-08-30T08:39:29.467469+00:00.