# google-research/albert

ALBERT: A Lite BERT for Self-supervised Learning of Language Representations

Repository: https://github.com/google-research/albert
Canonical: https://ross.abutalabs.com/products/google-research-albert
Language: Python
License: Apache-2.0
License Family: permissive
Archived: true
Last push: 2023-04-14T18:02:55+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2472, "days_push": 1237, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3278, forks 576 (observed 2026-08-28T04:07:53.517484+00:00)

## What it is
Official TensorFlow implementation and pretrained checkpoints of ALBERT, a lite version of BERT for self-supervised learning of language representations. It provides model code, training scripts, and TF-Hub models in multiple sizes and languages including Chinese.

## Use cases
- fine-tune a lightweight BERT-style model on GLUE tasks
- download pretrained ALBERT checkpoints for question answering
- train a language representation model with fewer parameters
- fine-tune ALBERT on SQuAD for extractive question answering
- use Chinese pretrained language models for NLP tasks
- run a colab tutorial for GLUE fine-tuning

## When to choose
- you need parameter-efficient pretrained transformer checkpoints in TensorFlow 1.x
- you want strong NLP baselines like SQuAD, MNLI, or RACE fine-tuning
- you need Chinese BERT-style pretrained models

## When to avoid
- you need PyTorch or modern Hugging Face ecosystem integration
- you require active maintenance or recent TensorFlow 2.x support
- you want the latest state-of-the-art language models rather than a 2019-era architecture

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, nlp, llm-training, deep-learning
- domain: machine-learning, deep-learning
- platform: python
- tags: bert, language-models, tensorflow, pretrained-models, transfer-learning, google-research, natural-language-processing, linux, gpu

## Member repositories
- google-research/albert (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:53.517484+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 2026-08-29T18:43:21.204644+00:00, confidence not recorded.
  - readme: https://github.com/google-research/albert (fetched 2026-08-28T04:07:53.517484+00:00, sha 5e0ba9acc58b)
- Data as of 2026-08-30T08:39:29.467469+00:00.