MACE MP 0 'medium' foundation model for atomistic materials chemistry v000

MACE-MP-0 (Medium) model, as described in Batatia et al. (2024). Published at: https://github.com/ACEsuit/mace-foundations/releases/tag/mace_mp_0. This variant utilizes a maximal message equivariance of $L=1$, offering an optimal trade-off between computational cost and accuracy. It is a Message Passing Neural Network (MPNN) trained on over 150,000 inorganic crystals, capable of zero-shot prediction for a wide range of chemical systems. Batatia, I., et al. 'A foundation model for atomistic materials chemistry.' arXiv preprint arXiv:2401.00096 (2024). Architecture: 2 layers, 128 channels, $l_{max}=3$, 4-body messages. Parameters: 4,688,656