meta-pytorch/torchstore

The high-performance distributed tensor layer — load once, share everywhere.

Simple, safe way to store and distribute tensors

PyTorch Single Controller

A dict[str, torch.Tensor] that behaves like one tensor, and TensorClass: typed, batched dataclasses that compile.

A performant, memory-efficient checkpointing library for PyTorch applications, designed with large, complex distributed workloads in mind.

High-throughput tensor loading for PyTorch

Reverb is an efficient and easy-to-use data storage and transport system designed for machine learning research

TensorLy-Torch: Deep Tensor Learning with TensorLy and PyTorch

PyTorch-native post-training at scale