PyTorch Lightning Tutorial: Using TorchMetrics and Lightning Flash

TorchMetrics unsurprisingly provides a modular approach to define and track useful metrics across batches and devices, while Lightning Flash offers a suite of functionality facilitating more efficient transfer learning and data handling, and a recipe book of state-of-the-art approaches to typical deep learning problems.

We’ll start by adding a few useful classification metrics to the MNIST example we started with earlier. We’ll also swap out the PyTorch Lightning Trainer object with a Flash Trainer object, which will make it easier to perform transfer learning on a new classification problem. We’ll then train our classifier on a new dataset, CIFAR10, which we’ll use as the basis for a transfer learning example to CIFAR100.

from DZone.com Feed https://ift.tt/3mZd2zy