# CLUEbenchmark/CLUE

中文语言理解测评基准 Chinese Language Understanding Evaluation Benchmark: datasets, baselines, pre-trained models, corpus and leaderboard

Repository: https://github.com/CLUEbenchmark/CLUE
Canonical: https://ross.abutalabs.com/products/clue
Homepage: http://www.CLUEbenchmarks.com
Language: Python
License Family: other
Topics: nlu, benchmark, chinese, corpus, dataset, bert, albert, chineseglue, glue, roberta, language-model, pretrained-models, transformers, tensorflow, pytorch
Last push: 2026-02-06T11:41:17+00:00

## Health v2 (maintenance only)
Score: 62/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 66, release rhythm 35, longevity 100
- inputs: {"age_days": 2476, "days_push": 208, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4279, forks 544 (observed 2026-08-28T04:08:41.129971+00:00)

## What it is
CLUE is the Chinese Language Understanding Evaluation Benchmark, providing representative datasets, baseline and pre-trained models, corpora, and a public leaderboard for Chinese NLP tasks. It also encompasses SuperCLUE, an evolving suite of benchmarks for evaluating Chinese large language models across reasoning, agents, safety, and multimodal capabilities.

## Use cases
- evaluate Chinese language understanding models on standard tasks
- compare pretrained models like BERT, RoBERTa, and ALBERT on Chinese NLU
- find Chinese NLP datasets for classification, NER, and reading comprehension
- benchmark Chinese large language models against a leaderboard
- download Chinese corpora for pretraining or language modeling
- reproduce baseline results for Chinese text classification tasks

## When to choose
- you need standardized Chinese NLU evaluation datasets and baselines
- you want to compare your model against published Chinese leaderboard scores
- you need large Chinese corpora for pretraining or domain adaptation
- you are evaluating Chinese LLMs with SuperCLUE benchmarks

## When to avoid
- your task is English-only NLU evaluation (use GLUE or SuperGLUE instead)
- you need a maintained software library rather than datasets and benchmarks
- you require a permissively licensed codebase, as the repository has no license

## Facets
- artifact type: dataset
- maturity: active
- function: machine-learning, nlp, benchmarking, data-science
- domain: machine-learning, large-language-models, artificial-intelligence
- platform: python, cross-platform
- tags: chinese-nlp, benchmark, leaderboard, pretrained-models, corpus, nlu, bert, evaluation, natural-language-processing

## Member repositories
- CLUEbenchmark/CLUE (main) score 62

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:41.129971+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 2026-08-29T18:21:58.545043+00:00, confidence not recorded.
  - readme: https://github.com/CLUEbenchmark/CLUE (fetched 2026-08-28T04:08:41.129971+00:00, sha 0e201a6c31c7)
  - homepage: http://www.CLUEbenchmarks.com (fetched 2026-08-29T09:11:23.898650+00:00, sha 5e0ce25bccd2)
  - site_page: https://www.cluebenchmarks.com:443/aboutClue.html (fetched 2026-08-29T09:11:23.901875+00:00, sha b318c2db06a4)
- Data as of 2026-08-30T08:39:29.467469+00:00.