# Lightning-Universe/lightning-flash

Your PyTorch AI Factory - Flash enables you to easily configure and run complex AI recipes for over 15 tasks across 7 data domains

Repository: https://github.com/Lightning-Universe/lightning-flash
Canonical: https://ross.abutalabs.com/products/lightning-flash
Homepage: https://lightning-flash.readthedocs.io
Language: Python
License: Apache-2.0
License Family: permissive
Topics: pytorch-lightning, deep-learning, machine-learning, pytorch, tasks-flash, classification, tabular-data, object-detection, icevision, open3d, fiftyone, torch-geometric, pytorch-video
Archived: true
Last push: 2023-10-08T14:28:44+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2043, "days_push": 1060, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1722, forks 211 (observed 2026-08-28T04:05:27.723047+00:00)

## What it is
Lightning Flash is a high-level PyTorch library built on PyTorch Lightning that provides ready-made 'recipes' for over 15 AI tasks across 7 data domains, including image classification, object detection, segmentation, tabular data, and more. It lets users load data, configure a task, and train or fine-tune state-of-the-art models with just a few lines of code.

## Use cases
- fine-tune an image classifier on my own photos with minimal code
- train an object detection model without writing training loops
- do semantic segmentation on a folder of images and masks
- build a text classifier using pretrained transformers
- train a model on tabular csv data quickly
- quickly prototype deep learning models across many task types
- serve a trained model as an API endpoint

## When to choose
- you want fast, few-lines-of-code fine-tuning of pretrained models for common tasks
- you are using PyTorch Lightning and want higher-level task abstractions
- you need to cover many data modalities (image, text, tabular, video, audio, graph) with one consistent API
- you are prototyping or teaching deep learning without boilerplate

## When to avoid
- you need full control over model architecture and training loops
- you need cutting-edge or recently released models and techniques
- you require a project with frequent updates and active maintenance
- your task is not among the supported recipes

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, image-processing, nlp, audio-processing, video-processing, data-science
- domain: machine-learning, deep-learning, computer-vision, developer-tools
- platform: python, cross-platform
- tags: pytorch, pytorch-lightning, transfer-learning, fine-tuning, pretrained-models, image-classification, object-detection, semantic-segmentation, tabular-data, high-level-api, natural-language-processing

## Member repositories
- Lightning-Universe/lightning-flash (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:27.723047+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 2026-08-30T03:33:58.575869+00:00, confidence not recorded.
  - readme: https://github.com/Lightning-Universe/lightning-flash (fetched 2026-08-28T04:05:27.723047+00:00, sha 2354653d31a8)
- Data as of 2026-08-30T08:39:29.467469+00:00.