SageMakerCV offers a collection of deep learning computer vision models in both PyTorch and TensorFlow. These models incorporate custom CUDA operations, SageMaker Distributed Data Parallel for distributed training, mixed precision training options, and algorithmic improvements.
For more information see the SageMakerCV blog, as well as the Mask RCNN in under 7 minutes blog from Re:Invent 2020.
Tutorial notebooks to get started in PyTorch and Tensorflow are provided in their respective directories.
See CONTRIBUTING for more information.
This project is licensed under the Apache-2.0 License.