v0.3.0

[ilyes319]

This Release Introduces:

1. Committee of models in the ASE calculator.
2. Active learning for MD in the ASE script.
3. Implementation of the Huber loss function.
4. Can use dipole model within the ASE calculator.
5. A LAMMPS wrapper for executing LAMMPS MD simulations.
6. The matscipy neighbor list set as the default.
7. Gaussian radial basis functions.
8. Element-dependent radial basis functions using a MLP on the radial components and node features of the sender and receiver.
9. MACE descriptors from the ASE calculator.
10. A universal pretrained MACE model on the Materials Project, available as an ASE calculator.

Major Interface Changes:

1. The scripts have been moved to mace/cli/ from the previous scripts directory to facilitate access.
2. The mace_run_train command is now available if MACE is installed via pip.

What's Changed

• merge readme edits to develop by @davkovacs in merge readme edits to develop #76
• Require stress in shape (3,3) or (6,) by @LarsSchaaf in Require stress in shape (3,3) or (6,) #77
• Committee error by @LarsSchaaf in Committee error #78
• Doc update by @ilyes319 in Doc update #85
• no gradient with respect to model params in eval by @davkovacs in no gradient with respect to model params in eval #83
• Update active_learning_md.py by @sandipde in Update active_learning_md.py #89
• Update utils.py by @sandipde in Update utils.py #91
• Huberloss by @ilyes319 in Huberloss #96
• not break if exponential LR by @bigd4 in not break if exponential LR #98
• hopefully limit unnecessary memory use. by @wcwitt in hopefully limit unnecessary memory use. #103
• ASE calculator refactor and jit dipole models by @davkovacs in ASE calculator refactor and jit dipole models #95
• Fix unweighted EnergyForcesLoss by @chaitjo in Fix unweighted EnergyForcesLoss #100
• Lammps wrapper by @wcwitt in Lammps wrapper #108
• Update Installation by @ilyes319 in Update Installation #129
• matscipy neighbour list as default by @ilyes319 in matscipy neighbour list as default #116
• Gaussian radial basis and element dependent radial by @ilyes319 in Gaussian radial basis and element dependent radial #131
• MACE descriptors by @ilyes319 in MACE descriptors #144
• Add readme section 'Pretrained Universal MACE Checkpoints' by @janosh in Add readme section 'Pretrained Universal MACE Checkpoints' #188
• fix table of content readme by @ilyes319 in fix table of content readme #190
• Scripts entrypoints by @bernstei in Scripts entrypoints #178
• add foundations model calculators by @ilyes319 in add foundations model calculators #204
• add deprecated model_path arg to asse calculator by @davkovacs in add deprecated model_path arg to asse calculator #209
• MACE 0.3.0 release by @ilyes319 in MACE 0.3.0 release #142

New Contributors

• @LarsSchaaf made their first contribution in Require stress in shape (3,3) or (6,) #77
• @sandipde made their first contribution in Update active_learning_md.py #89
• @bigd4 made their first contribution in not break if exponential LR #98
• @chaitjo made their first contribution in Fix unweighted EnergyForcesLoss #100

Full Changelog: v0.2.0...v0.3.0

---

This discussion was created from the release v0.3.0.

[chaitjo]

Congrats! I'm curious as to how impactful was switching to gaussian RBF vs. the previous RBF function in MACE?