# fastai/imagenette

A smaller subset of 10 easily classified classes from Imagenet, and a little more French

Repository: https://github.com/fastai/imagenette
Canonical: https://ross.abutalabs.com/products/imagenette
Language: Jupyter Notebook
License: Apache-2.0
License Family: permissive
Last push: 2022-09-26T22:55:07+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2738, "days_push": 1437, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1131, forks 78 (observed 2026-08-28T04:03:42.479044+00:00)

## What it is
Imagenette is a subset of 10 easily classified ImageNet classes, with companion datasets Imagewoof (10 hard dog breeds) and Image网, provided in full-size, 320px, and 160px variants. It includes CSV files with 1-50% noisy labels and a 70/30 train/validation split, aimed at fast computer vision experimentation and benchmarking.

## Use cases
- quickly test image classification models on a small imagenet subset
- benchmark computer vision training recipes without full imagenet
- experiment with learning under label noise using noisy label CSVs
- teach deep learning image classification with a lightweight dataset
- compare model performance on easy (imagenette) vs hard (imagewoof) classes

## When to choose
- you need a small, fast image classification dataset for prototyping or education
- you want to benchmark vision models without downloading full ImageNet
- you want to study robustness to label noise

## When to avoid
- you need the full 1000-class ImageNet dataset
- you need datasets outside image classification
- you need actively maintained tooling rather than static dataset downloads

## Facets
- artifact type: dataset
- maturity: stable
- function: machine-learning, image-processing
- domain: computer-vision, image-processing, machine-learning
- platform: cross-platform
- tags: image-classification, imagenet-subset, computer-vision-benchmark, noisy-labels, fastai

## Member repositories
- fastai/imagenette (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:42.479044+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 2026-08-30T06:38:02.297164+00:00, confidence not recorded.
  - readme: https://github.com/fastai/imagenette (fetched 2026-08-28T04:03:42.479044+00:00, sha e410a6dee2c2)
- Data as of 2026-08-30T08:39:29.467469+00:00.