# ymcui/Chinese-ELECTRA

Pre-trained Chinese ELECTRA（中文ELECTRA预训练模型）

Repository: https://github.com/ymcui/Chinese-ELECTRA
Canonical: https://ross.abutalabs.com/products/chinese-electra
Homepage: http://electra.hfl-rc.com
Language: Python
License: Apache-2.0
License Family: permissive
Topics: nlp, bert, electra, pre-trained-model, chinese, chinese-electra, language-model, pytorch, tensorflow
Last push: 2026-04-19T00:57:39+00:00

## Health v2 (maintenance only)
Score: 67/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 78, release rhythm 35, longevity 100
- inputs: {"age_days": 2354, "days_push": 137, "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 1434, forks 165 (observed 2026-08-28T04:04:43.325940+00:00)

## What it is
Chinese-ELECTRA is a collection of pre-trained Chinese ELECTRA language models released by the HIT-iFLYTEK joint lab, built on the official ELECTRA training code and large-scale Chinese corpora. It provides downloadable checkpoints (small, base, large, legal-domain variants) loadable via HuggingFace Transformers and PaddleHub for downstream Chinese NLP tasks.

## Use cases
- fine-tune a Chinese language model for text classification
- Chinese question answering and reading comprehension models
- named entity recognition on Chinese text
- use a lightweight Chinese model with BERT-base-level performance
- legal-domain Chinese NLP with pre-trained checkpoints
- load a Chinese ELECTRA model with HuggingFace Transformers

## When to choose
- you need pre-trained Chinese transformer checkpoints for downstream NLP tasks
- you want ELECTRA's efficiency, e.g. ELECTRA-small matching BERT-base with 1/10 the parameters
- you need domain-specific Chinese models such as legal-domain variants
- you work in PyTorch, TensorFlow, or PaddlePaddle ecosystems

## When to avoid
- you need generative or instruction-following LLMs rather than encoder models
- your target language is not Chinese
- you need a ready-made application rather than model weights and training code

## Facets
- artifact type: library
- maturity: stable
- function: machine-learning, nlp
- domain: large-language-models, machine-learning
- platform: python, cross-platform
- tags: pre-trained-model, electra, chinese-nlp, bert, transformers, huggingface, pytorch, tensorflow, natural-language-processing

## Member repositories
- ymcui/Chinese-ELECTRA (main) score 67

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:43.325940+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 2026-08-30T04:36:56.297806+00:00, confidence not recorded.
  - readme: https://github.com/ymcui/Chinese-ELECTRA (fetched 2026-08-28T04:04:43.325940+00:00, sha 096eaa8b49be)
- Data as of 2026-08-30T08:39:29.467469+00:00.