MosaicML Streaming is an open-source project on GitHub with about 1,570 stars, most recently updated in 2026. Streaming is a data streaming library from MosaicML, now part of Databricks, for efficient neural network training. It streams training datasets from cloud or local storage on demand, with deterministic sample order, shuffling and mid-epoch resumption, so large-scale PyTorch training is never bottlenecked by data loading.
Published on GitHub under the Apache-2.0 license, MosaicML Streaming has attracted a community of about 1,570 stars, which places it among the more visible open-source AI projects in its niche. Development was still active as of 2026, based on the repository’s latest updates. Because the full source code is public, teams can audit how it works, adapt it to their own stack, and self-host it rather than depending on a closed hosted product.
The repository is tagged with topics including dataset, deep-learning, machine-learning, neural-network, pytorch, which gives a quick picture of where it fits in the open-source AI ecosystem. New users can explore the examples and documentation in the repository to understand the main workflows before integrating it into their own projects, and the public issue tracker shows what the maintainers are actively working on and what the community is asking for.
Streaming is Apache-2.0 open source and free to use in any training pipeline.
MosaicML Streaming is best suited for ML platform and training engineers running large model training jobs who need fast, reliable data delivery. Anyone evaluating it should start with the GitHub repository, where the documentation, examples and issue tracker show exactly what the current version can do.