Deci Unveils a Deep Learning Training Library for Developers

Deci, a deep learning development company that harnesses artificial intelligence (AI) to build artificial intelligence, is releasing SuperGradients, an “all-in-one” deep learning training library for computer vision models.

SuperGradients allows developers to train PyTorch-based models for the most common computer vision tasks, including object detection, image classification, and semantic segmentation using just one training script.

In addition, users can easily load and adjust the latest pre-trained models (YOLOv5, DDRNet, EfficientNet, RegNet, ResNet, MobileNet, etc.), many of which are optimized to offer higher accuracy compared to existing training libraries, according to the vendor.

Developed by Deci’s deep learning experts, SuperGradients offers a wide range of pre-trained, production-ready deep learning models that have been tested in production environments (i.e. made into optimized, deployable runtime implementers such as OpenVINO, TensorRT, and ONNX). The library also includes proven training recipes to easily reproduce training results, making artificial intelligence accessible to everyone.

“As deep learning matures and is widely adopted, model development and training processes must be streamlined. In order to fulfill the promise of deep learning, teams require tools that can facilitate the engineering selection phase, as well as enable them to achieve better training outcomes faster,” said Yonatan Geoffman, Partner Founder and CEO of Desi.” This is why we decided to release SuperGradients as an open source solution for the entire AI community to take advantage of. By offering developers better tools for building, improving, and deploying models, we help simplify the entire deep learning lifecycle and enable developers to focus on what they do best — creating innovative AI solutions to solve the world’s most complex problems. “

SuperGradients benefits include:

  • Easy access to training modules and fine-tuning for any computer vision task with just one training repository.
  • Training recipes that outperform other training libraries, as well as code examples for a wide range of computer vision models created by DCI’s team of deep learning experts, thus eliminating uncertainty and shortening the training lifecycle.
  • Plug and play hyperparameters for training, dataset and architecture
  • Seamless integration of deep learning models into users’ code base is hassle-free.

SuperGradients also supports fine-tuning and easy retraining of DeciNets, a set of state-of-the-art deep learning models built with Deci’s Automated Neural Architecture (AutoNAC) technology. DeciNets is currently used by leading AI teams around the world for multiple use cases.

For more information on this news, visit

Leave a Comment