generated: '2026-07-20'
method: searched
source: https://pypi.org/user/mosaicml/ ; https://github.com/mosaicml
packages:
  - language: python
    registry: pypi
    name: mosaicml
    url: https://pypi.org/project/mosaicml/
    install: pip install mosaicml
    version: 0.32.1
    official: true
    summary: >-
      Composer - a PyTorch library that enables you to train neural networks
      faster, at lower cost, and to higher accuracy.
    source: https://github.com/mosaicml/composer
  - language: python
    registry: pypi
    name: llm-foundry
    url: https://pypi.org/project/llm-foundry/
    install: pip install llm-foundry
    version: 0.22.0
    official: true
    summary: >-
      LLM Foundry - code for training, finetuning, evaluating and deploying
      large language models with Composer and the MosaicML platform.
    source: https://github.com/mosaicml/llm-foundry
  - language: python
    registry: pypi
    name: mosaicml-streaming
    url: https://pypi.org/project/mosaicml-streaming/
    install: pip install mosaicml-streaming
    version: 0.13.0
    official: true
    summary: >-
      Streaming - create PyTorch compatible datasets that can be streamed from
      cloud-based object stores.
    source: https://github.com/mosaicml/streaming
  - language: python
    registry: pypi
    name: mosaicml-cli
    url: https://pypi.org/project/mosaicml-cli/
    install: pip install mosaicml-cli
    version: 0.8.3
    official: true
    imports_as: mcli
    summary: >-
      Interact with Databricks Mosaic AI training from Python or a command-line
      interface (the mcli tool).
    source: https://github.com/mosaicml/mosaicml-cli
notes: >-
  All packages are published by the mosaicml org on PyPI and GitHub and are
  first-party. The unscoped PyPI name "mcli" is an UNRELATED third-party project
  (Monaize MarketMaker CLI) and is deliberately excluded. No first-party npm,
  Maven, NuGet, RubyGems, Go-module, Packagist, or crates.io packages were found.