Intel oneAPI AI Analytics Toolkit

Optimized Deep Learning Training and Inference

Leverage popular, Intel-optimized frameworks—TensorFlow and PyTorch—to use the full power of Intel® architecture and yield high performance for training and inference. The toolkit also provides open-source, pretrained, machine-learning models to expedite development along with the Intel® Neural Compressor for automatic accuracy-driven tuning strategies and additional model optimizations for performance, model size and memory footprint.

Drop-in Acceleration for Classical ML Algorithms

Increase machine-learning model accuracy and performance for popular algorithms in Scikit-Learn and XGBoost on Intel architectures, with simple one-line of code change. The toolkit further helps to scale out efficiently to clusters and perform distributed machine learning with big data.

Simplified Scaling across Multi-node Pandas DataFrames

Seamlessly scale and accelerate Pandas workflows to multicores and multi-nodes with only one line of code change using the Intel® Distribution of Modin*, an extremely lightweight and performant parallel DataFrame system.