LBANN (Livermore Big Artificial Neural Network Toolkit)
Project record
Maintained by Lawrence Livermore National Laboratory (LBANN project)516
LBANN is an open-source deep learning training framework from Lawrence Livermore National Laboratory, built for high-performance computing systems. Its documentation describes combining model parallelism through domain decomposition with data parallelism and ensemble training, and support for supervised, self-supervised, unsupervised, and adversarial (GAN) training. The repository's main branch now holds LBANNv2, a pre-alpha Python package described as LBANN's core integration with PyTorch; the earlier toolkit remains on the v1.x branches.6324
- Repository: Repository (LBANN/lbann) (external site: github.com)
- Documentation: Documentation (external site: lbann.readthedocs.io)
- License: LICENSE (Apache 2.0) (external site: github.com)
- Release notes: Releases (external site: github.com)
Availability and license
Component reuse rights
- weights
- Unknown — no complete fact-level rights review
- code
- Reviewed qualifying license recorded — check scope and conditions
- data
- Unknown — no complete fact-level rights review
- documentation
- Unknown — no complete fact-level rights review
No complete system-rights review is recorded for this release.
Public materials checklist
| Item | Status | Notes and evidence |
|---|---|---|
| Source codeIs the source code publicly readable? | Public | Public GitHub repository under the Apache License 2.0.15 |
| DocumentationIs user documentation published? | Public | Read the Docs pages cover building and installing, running LBANN, examples, and a list of publications. These pages describe the v1.x toolkit; the main-branch README for LBANNv2 gives only brief build steps.62 |
| InstallationAre installation instructions or packages publicly available? | Public | The v1.x README names Spack ("spack install lbann") as the preferred install method. The LBANNv2 README says to install PyTorch first and then install the package with pip from a clone.32 |
| Supported platformsAre supported operating systems or hardware documented? | Partial | The v1.x README says LBANN is optimized for one GPU per MPI rank, and the documentation has build guides for known HPC centers and Livermore Computing systems. The LBANNv2 package metadata lists Python 3.9 through 3.13. No full hardware or operating-system support matrix was found.364 |
| Release statusAre versioned releases published? | Public | Versioned GitHub releases run through v0.104, published November 8, 2023. The LBANNv2 package on the main branch is versioned 0.0.1 and marked "Pre-Alpha" in its metadata.74 |
What it is useful for
Run and use notes
- The v1.x README documents running LBANN through an MPI launcher with model, optimizer, and data-reader configuration files in prototext format, and recommends assigning one GPU per MPI rank.3
Organization context
U.S. eligibility
Eligible · basis: U.S.-governed project
LBANN's license file assigns copyright to Lawrence Livermore National Security, LLC and other LBANN project developers, states that the code was produced at Lawrence Livermore National Laboratory, and carries the LLNL code identifier LLNL-CODE-697807. LLNS operates LLNL, a U.S. Department of Energy national laboratory in Livermore, California, for DOE's National Nuclear Security Administration. The repository now sits in the LBANN GitHub organization; the license names both that organization and the LLNL organization as project locations.518
Sources
This listing is not an endorsement, a safety assessment, or a federal approval.