See how to use Intel® Extension for PyTorch* for training and inference on the MedMNIST datasets. These datasets are a collection of 10 MNIST-like open datasets on various medical-imaging classification tasks, such as pathology, chest X-ray, and optical coherence tomography (OCT) images. The demonstration runs on Intel® Tiber™ Developer Cloud. It is compared against stock PyTorch and shows the performance gain that Intel Extension for PyTorch offers.

Séverine Habert is a deep learning software engineer at Intel who helps data scientists use AI Tools. She holds a PhD in medical imaging from Technical University of Munich.