# google-research/bert

TensorFlow code and pre-trained models for BERT

Repository: https://github.com/google-research/bert
Canonical: https://ross.abutalabs.com/products/bert
Homepage: https://arxiv.org/abs/1810.04805
Language: Python
License: Apache-2.0
License Family: permissive
Topics: nlp, google, natural-language-processing, natural-language-understanding, tensorflow
Archived: true
Last push: 2024-07-23T23:39:41+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": 2869, "days_push": 771, "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 40046, forks 9678 (observed 2026-08-28T04:12:08.099905+00:00)

## What it is
Google Research's official TensorFlow implementation of BERT, the Bidirectional Encoder Representations from Transformers language model, along with pre-trained model checkpoints. It provides code for pre-training and fine-tuning BERT on NLP tasks like question answering, sentence classification, and language inference.

## Use cases
- fine-tune BERT for text classification
- download pre-trained BERT checkpoints
- train a question answering model on SQuAD
- run NLP experiments with TensorFlow
- use small BERT models for limited compute
- pre-train a transformer language model from scratch

## When to choose
- you need the original reference BERT implementation in TensorFlow
- you want official pre-trained checkpoints including compact models
- you are reproducing the BERT paper's GLUE or SQuAD results

## When to avoid
- you prefer PyTorch or Hugging Face Transformers ecosystem
- you need generative or decoder-based LLMs
- you want actively developed features and recent model architectures

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, nlp, deep-learning, transformers
- domain: machine-learning, deep-learning
- platform: python
- tags: bert, pretrained-models, tensorflow, language-model, fine-tuning, google-research, natural-language-processing, gpu

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

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:12:08.099905+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 2026-08-29T16:24:05.394544+00:00, confidence not recorded.
  - readme: https://github.com/google-research/bert (fetched 2026-08-28T04:12:08.099905+00:00, sha 9f0784648b42)
  - homepage: https://arxiv.org/abs/1810.04805 (fetched 2026-08-29T07:46:42.202541+00:00, sha cd9c49f5e17f)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T07:46:42.211782+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T07:46:42.216694+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T07:46:42.218913+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T07:46:42.214306+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.