# brightmart/albert_zh

A LITE BERT FOR SELF-SUPERVISED LEARNING OF LANGUAGE REPRESENTATIONS, 海量中文预训练ALBERT模型

Repository: https://github.com/brightmart/albert_zh
Canonical: https://ross.abutalabs.com/products/albert_zh
Homepage: https://arxiv.org/pdf/1909.11942.pdf
Language: Python
License Family: other
Topics: albert, bert, roberta, xlnet, tensorflow, pytorch, pre-trained-model, chinese-corpus, pre-trained
Last push: 2022-11-21T08:02:01+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": 2533, "days_push": 1381, "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 3982, forks 743 (observed 2026-08-28T04:08:31.688558+00:00)

## What it is
A repository providing pre-trained ALBERT models for Chinese language, implemented in TensorFlow with PyTorch and Keras conversions. It includes lightweight variants like albert_tiny that retain most of BERT's accuracy with a fraction of the parameters, plus scripts for fine-tuning on Chinese NLP tasks.

## Use cases
- pretrain or fine-tune a small Chinese language model
- run Chinese text classification with a lightweight BERT alternative
- compute Chinese sentence semantic similarity
- deploy a compact NLP model on mobile via TensorFlow Lite
- benchmark Chinese models on CLUE tasks

## When to choose
- you need a small, fast pre-trained model for Chinese NLP tasks
- you want TensorFlow or PyTorch checkpoints for ALBERT in Chinese
- low-latency or on-device inference matters more than peak accuracy

## When to avoid
- you need multilingual or English-only models
- you want the latest transformer architectures or active community support
- you require a permissively licensed project - no license is specified

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, nlp, llm-training, transformers
- domain: machine-learning, large-language-models
- platform: python
- tags: albert, pre-trained-models, chinese-nlp, tensorflow, pytorch, model-zoo, fine-tuning, natural-language-processing, linux, gpu

## Member repositories
- brightmart/albert_zh (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:31.688558+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 2026-08-29T18:24:16.024390+00:00, confidence not recorded.
  - readme: https://github.com/brightmart/albert_zh (fetched 2026-08-28T04:08:31.688558+00:00, sha 8fa4830a1263)
- Data as of 2026-08-30T08:39:29.467469+00:00.