fairchem-core-2.13.0

What’s Changed

Lots of Parallel and Batch inference upgrades:

• Added new highly efficient batch inference integration with Ray
• Graph building is now parallelized, UMA-s runs real MD with 1 ns/per day on 100k+ atoms with GP

New Features / Enhancements

• Early Partition and Parallel Graph Gen (#1630) @rayg1234
• Make a checkpointable Ray launcher (#1666) @rayg1234
• Batch server unit (#1622) @lbluque
• Formation energy calculator (#1538) @lbluque

Documentation

• UMA catalysis tutorial (#1667) @zulissimeta

Dependencies

• Add ray as dependency (#1675) @lbluque
• upgrade numpy/numba (#1537) @lbluque
• allow newer lmdb versions (#1655) @lbluque

fairchem-core-2.12.0

What’s Changed

New Features / Enhancements

• merge moe head with test (#1647) @misko
• Allow CPU graph generation in calculator (#1628) @lbluque
• remove extra assert; blocks edge_chunk_size!=None (#1633) @misko
• padded edges (#1621) @misko
• Add job names to ray cluster slurm jobs (#1625) @rayg1234

Bug Fixes

• Fix bug breaking mult-node ray runs (#1635) @rayg1234

fairchem_data_omc-0.1.0

What’s Changed

New Features / Enhancements

• OMC25 dataset release (#1392) @gvahe

Documentation

• OMC25 dataset release (#1392) @gvahe

Dependencies

• OMC25 dataset release (#1392) @gvahe

fairchem_data_omat-0.2.0

What’s Changed

New Features / Enhancements

• add compatibility corrections (#1603) @lbluque

Documentation

• Update OMat24 README.md (#1560) @lbluque

fairchem-core-2.11.0

What’s Changed

Lots of improvements running Parallel UMA!

New Features / Enhancements

• Update only gputils (splitting previous PR) (#1592) @misko
• Allow merged MoLE to predict different sized cells with same composition (#1620) @lbluque
• Use NCCL with ParallelMLIPPredictUnitRay when possible (#1587) @misko
• Clamp rotation - fine/tuning unblock (#1574) @misko

Bug Fixes

• fixed batched parallel predict unit (#1606) @misko
• Pin jupyter-book (#1609) @rayg1234

Dependencies

• Add tests for python<=3.13 (#1600) @zulissimeta

fairchem-lammps-0.3.0

What's Changed

New Features

• Refactor launchers to add ray cluster mode by @rayg1234 in #1531
• fix_checkpoint_migration_for_non_uma by @misko in #1508
• Data OMat by @lbluque in #1520
• Enable multinode Ray by @rayg1234 in #1526
• Fix ORCA calculation script reference in omol25.md by @Andrew-S-Rosen in #1545
• Improve multi gpu docs and interface by @rayg1234 in #1580

Bug Fixes

• Use str ids in AseReadDataset by @lbluque in #1555
• Fix virial sign in NPT ensemble, wrap coordinates to unit cell, and support charge/spin configuration by @yu9824 in #1546
• Fix seeds for predict test by @rayg1234 in #1584
• Fix build yaml by @rayg1234 in #1590

Documentation

• Rename ParallelMLIPPredictUnit by @rayg1234 in #1582

Other Changes

• Dependabot config by @lbluque in #1557
• Bump actions/setup-python from 5 to 6 by @dependabot[bot] in #1471
• Revert ipykernel to fix build docs by @zulissimeta in #1583

New Contributors

• @yu9824 made their first contribution in #1546

Full Changelog: fairchem_lammps-0.2.0...fairchem_lammps-0.3.0

fairchem_core-2.10.0

What's Changed

New Features

• Fix ORCA calculation script reference in omol25.md by @Andrew-S-Rosen in #1545
• Improve multi gpu docs and interface by @rayg1234 in #1580

Bug Fixes

• Fix virial sign in NPT ensemble, wrap coordinates to unit cell, and support charge/spin configuration by @yu9824 in #1546
• Fix seeds for predict test by @rayg1234 in #1584

Documentation

• Rename ParallelMLIPPredictUnit by @rayg1234 in #1582

Other Changes

• Bump actions/setup-python from 5 to 6 by @dependabot[bot] in #1471
• Revert ipykernel to fix build docs by @zulissimeta in #1583

fairchem_data_omat-0.1.0

What’s Changed

New Features / Enhancements

• Data OMat (#1520) @lbluque

fairchem-core-2.9.0

What’s Changed

New Features / Enhancements

• Refactor launchers to add ray cluster mode (#1531) @rayg1234

Bug Fixes

• Use str ids in AseReadDataset (#1555) @lbluque

fairchem-lammps-0.2.0

What’s Changed

Documentation

• Add training docs (#1506) @rayg1234

Deprecations

• remove wigner_cuda / cuda_graph (#1523) @misko