Welcome to SuperGradients, a free, open-source training library for PyTorch-based deep learning models. SuperGradients allows you to train or fine-tune SOTA pre-trained models for all the most commonly applied computer vision tasks with just one training library. We currently support object detection, image classification and semantic segmentation for videos and images.
Easily load and fine-tune production-ready, pre-trained SOTA models that incorporate best practices and validated hyper-parameters for achieving best-in-class accuracy (Yolox, PP-YoloE, STDC, DDRNet, and PP-LiteSeg).
Why do all the grind work, if we already did it for you? leverage tested and proven recipes & code examples for a wide range of computer vision models generated by our team of deep learning experts. Easily configure your own or use plug & play hyperparameters for training, dataset, and architecture.
All SuperGradients models’ are production ready in the sense that they are compatible with deployment tools such as TensorRT (Nvidia) and OpenVINO (Intel) and can be easily taken into production. With a few lines of code you can easily integrate the models into your codebase.
Check out our Quickstart tutorial to get learn the basic of SuperGradients.
You can also start from our tutorial on Detection, Segmentation or Pose Estimation.
Version 3.6.1 (March 6, 2024)
- A dependency from
pycocotools
has been removed from SG, we don't rely anymore on this package to parse COCO dataset json. - A
Trainer.ptq
andTrainer.qat
methods now allow granular control on for the model should be exported (with or without pre-/post-processing). - A
model.predict
now hasfp16
argument (Default isTrue
) which one can use to disable mixed precision feature (Addressing issues on GTX 16XX series) - Fixed a bug in missing min-max image normalization in
plot()
method for detection dataset. - Removed
deci-common
from[pro]
requirements. - Updated YoloNAS-Pose fine-tunining for Animals Pose Dataset notebook.
Version 3.6.0 (Jan 25, 2024)
- Added segmentation samples and support for albumentation transforms for segmentation
- Implemented distance-based detection matching in
DetectionMetrics
as an enhancement (by @DimaBir) - New training hyperparameter - finetune, and multiple LR assignment read about it https://github.com/Deci-AI/super-gradients/blob/master/documentation/source/LRAssignment.md
- Enhanced
ImagePermute
processing inclusion - Improved dataset plotting and plot functionality
- A new API for checking model input compatibility
- Extended
predict()
support for segmentation models
Version 3.5.0 (November 23, 2023)
- Support for long videos in
model.predict()
(by @hakuryuu96) - Added support for multiple test loaders in
train_from_config
- Added skip_resize to
model.predict()
to support large images and small objects
Version 3.4.0 (November 6, 2023)
- YoloNAS-Pose model released - a new frontier in pose estimation
- Added option to export a recipe to a single YAML file or to a standalone train.py file
- Other bugfixes & minor improvements. Full release notes available here
If you are using SuperGradients library in your research, please cite SuperGradients deep learning training library.
If you want to be a part of SuperGradients growing community, hear about all the exciting news and updates, need help, request for advanced features, or want to file a bug or issue report, we would love to welcome you aboard!
-
Slack is the place to be and ask questions about SuperGradients and get support. Click here to join our Slack
-
To report a bug, file an issue on GitHub.
-
Join the SG Newsletter for staying up to date with new features and models, important announcements, and upcoming events.
This project is released under the Apache 2.0 license.
@misc{supergradients,
doi = {10.5281/ZENODO.7789328},
url = {https://zenodo.org/record/7789328},
author = {Aharon, Shay and {Louis-Dupont} and {Ofri Masad} and Yurkova, Kate and {Lotem Fridman} and {Lkdci} and Khvedchenya, Eugene and Rubin, Ran and Bagrov, Natan and Tymchenko, Borys and Keren, Tomer and Zhilko, Alexander and {Eran-Deci}},
title = {Super-Gradients},
publisher = {GitHub},
journal = {GitHub repository},
year = {2021},
}