# google-research/big_transfer

Official repository for the "Big Transfer (BiT): General Visual Representation Learning" paper.

Repository: https://github.com/google-research/big_transfer
Canonical: https://ross.abutalabs.com/products/big_transfer
Homepage: https://arxiv.org/abs/1912.11370
Language: Python
License: Apache-2.0
License Family: permissive
Topics: deep-learning, convolutional-neural-networks, imagenet, tensorflow2, jax, pytorch, transfer-learning
Archived: true
Last push: 2024-07-30T21:21:39+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": 2305, "days_push": 764, "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 1541, forks 175 (observed 2026-08-28T04:05:00.652977+00:00)

## What it is
Official repository for the Big Transfer (BiT) paper, providing ResNet models pre-trained on ImageNet and ImageNet-21k for transfer learning. It includes fine-tuning code and pretrained weights in TensorFlow 2, PyTorch, and Jax/Flax.

## Use cases
- fine-tune a pretrained image classifier on my own small dataset
- classify images with only a few labeled examples per class
- use ImageNet-21k pretrained ResNet weights in PyTorch or TensorFlow
- benchmark transfer learning on VTAB datasets
- get strong visual representations for downstream vision tasks

## When to choose
- you need high-quality pretrained CNN weights for image classification with limited data
- you want fine-tuning code across TF2, PyTorch, and Jax
- you want to reproduce BiT paper results

## When to avoid
- you need transformer-based vision models like ViT or CLIP
- you need actively developed features or new architectures
- your task is not image classification or visual representation learning

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, image-processing, deep-learning
- domain: computer-vision, machine-learning, deep-learning
- platform: python
- tags: transfer-learning, pretrained-models, resnet, imagenet, tensorflow2, pytorch, jax, fine-tuning, computer-vision, gpu

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

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:00.652977+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 2026-08-30T04:30:47.652581+00:00, confidence not recorded.
  - readme: https://github.com/google-research/big_transfer (fetched 2026-08-28T04:05:00.652977+00:00, sha 30240c3c1f3a)
  - homepage: https://arxiv.org/abs/1912.11370 (fetched 2026-08-29T11:32:20.801408+00:00, sha e36ed3227301)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T11:32:21.248366+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T11:32:21.270196+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T11:32:21.272122+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T11:32:21.268176+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.