Torchmeta is a collection of extensions and data-loaders for few-shot learning and meta-learning in PyTorch. It includes popular meta-learning benchmarks like Omniglot, Mini-ImageNet, and Tiered-ImageNet, with a unified interface for both few-shot classification and regression problems. The library provides a MetaModule base class that simplifies creating gradient-based meta-learning models, fully compatible with standard PyTorch DataLoader and torchvision.