google-research/mixmatch

Health v2 · maintenance only

Flags: no_releases archived

Adoption not part of the score

1139 stars · 161 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Reference implementation of MixMatch, a holistic semi-supervised learning algorithm from a Google Research paper. It includes dataset preparation scripts and training code for CIFAR, SVHN, and STL-10 semi-supervised benchmarks.

Use cases

• reproduce MixMatch semi-supervised learning results
• train image classifiers with few labeled examples
• compare semi-supervised learning methods on CIFAR-10
• study label-consistency regularization with mixup
• run semi-supervised experiments on SVHN and STL-10

When to choose

• you want the official reference implementation of the MixMatch paper
• you need semi-supervised image classification benchmarks with prepared splits
• you are doing research on semi-supervised learning techniques

When to avoid

• you need a production-ready or actively maintained ML library
• you want a framework-agnostic or PyTorch implementation
• you need semi-supervised learning for non-image modalities

Facets

library · maturity maintenance

machine-learning deep-learning machine-learning deep-learning artificial-intelligence python semi-supervised-learning research-code tensorflow mixup data-augmentation paper-reproduction gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="google-research/mixmatch")