# Deci-AI/super-gradients

Easily train or fine-tune SOTA computer vision models with one open source training library. The home of Yolo-NAS.

Repository: https://github.com/Deci-AI/super-gradients
Canonical: https://ross.abutalabs.com/products/super-gradients
Homepage: https://www.supergradients.com
Language: Jupyter Notebook
License: Apache-2.0
License Family: permissive
Topics: dependency-graph
Last push: 2026-02-24T12:31:01+00:00

## Health v2 (maintenance only)
Score: 54/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 69, release rhythm 8, longevity 100
- inputs: {"age_days": 1739, "days_push": 190, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 5052, forks 592 (observed 2026-08-28T04:09:09.044445+00:00)

## What it is
SuperGradients is an open-source PyTorch-based training library for building, training, and fine-tuning state-of-the-art computer vision models, and is the home of the Yolo-NAS object detection model. It provides pretrained model checkpoints and recipes for classification, object detection, segmentation, and pose estimation tasks.

## Use cases
- train a custom object detection model on my own dataset
- fine-tune Yolo-NAS for detection
- fine-tune a pretrained image classification model
- train a semantic segmentation model in PyTorch
- run pose estimation with a pretrained model
- export a vision model for production deployment

## When to choose
- you want a single library covering classification, detection, segmentation, and pose estimation
- you want access to Yolo-NAS and other SOTA pretrained checkpoints
- you prefer recipe-driven PyTorch training with sensible defaults

## When to avoid
- you need tasks outside computer vision such as NLP or audio
- you want a lightweight inference-only runtime rather than a training library
- you need a framework-agnostic solution beyond PyTorch

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, computer-vision, image-processing
- domain: computer-vision, deep-learning, machine-learning, image-processing
- platform: python, cross-platform
- tags: yolo-nas, object-detection, image-classification, semantic-segmentation, pose-estimation, model-training, fine-tuning, pretrained-models, pytorch, gpu

## Member repositories
- Deci-AI/super-gradients (main) score 54

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:09.044445+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 2026-08-29T18:17:20.919052+00:00, confidence not recorded.
  - readme: https://github.com/Deci-AI/super-gradients (fetched 2026-08-28T04:09:09.044445+00:00, sha b9ccda88fc8a)
  - homepage: https://www.supergradients.com (fetched 2026-08-29T08:57:53.615636+00:00, sha cb5e8223241b)
  - site_page: https://www.nvidia.com/en-eu/gtc/pricing?nvid=nv-bnr-659003 (fetched 2026-08-29T08:57:53.665706+00:00, sha b1be635140a0)
  - site_page: https://nvidianews.nvidia.com/news/nvidia-releases-vera-rubin-dsx-ai-factory-reference-design-and-omniverse-dsx-digital-twin-blueprint-with-broad-industry-support (fetched 2026-08-29T08:57:53.728951+00:00, sha e0add11ae768)
- Data as of 2026-08-30T08:39:29.467469+00:00.