# Tencent/tencent-ml-images

Largest multi-label image database; ResNet-101 model; 80.73% top-1 acc on ImageNet

Repository: https://github.com/Tencent/tencent-ml-images
Canonical: https://ross.abutalabs.com/products/tencent-ml-images
Language: Python
License: NOASSERTION
License Family: other
Topics: database, deep-learning, computer-vision
Last push: 2022-04-20T07:00:12+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2880, "days_push": 1596, "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 3063, forks 503 (observed 2026-08-28T04:07:41.219418+00:00)

## What it is
Tencent ML-Images is the largest open-source multi-label image database, containing ~17.6M training and ~88.7K validation image URLs annotated across 11,166 categories. It also provides a ResNet-101 model pre-trained on this database that achieves 80.73% top-1 accuracy on ImageNet via transfer learning.

## Use cases
- pretraining image classification models on a large multi-label dataset
- transfer learning for single-label image classification
- extracting image features with a pretrained ResNet-101
- training deep learning models on 11,166 image categories
- downloading large-scale image datasets from ImageNet and Open Images

## When to choose
- you need a very large multi-label image dataset for pretraining
- you want a strong pretrained ResNet-101 checkpoint for transfer learning
- your research requires broad category coverage beyond ImageNet-1k

## When to avoid
- you need a maintained pipeline on modern TensorFlow or PyTorch versions
- you cannot host or download tens of millions of images
- you need a small curated dataset for quick experiments

## Facets
- artifact type: dataset
- maturity: maintenance
- function: machine-learning, computer-vision, image-processing, deep-learning
- domain: computer-vision, image-processing, machine-learning, deep-learning
- platform: python
- tags: image-dataset, multi-label-classification, resnet-101, tensorflow, imagenet, pretrained-model, transfer-learning, linux

## Member repositories
- Tencent/tencent-ml-images (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:41.219418+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 2026-08-29T18:46:38.927899+00:00, confidence not recorded.
  - readme: https://github.com/Tencent/tencent-ml-images (fetched 2026-08-28T04:07:41.219418+00:00, sha 4fddf53ddbef)
- Data as of 2026-08-30T08:39:29.467469+00:00.