Automatically Fetching Public Pre-trained Mace Models in Mace Interface via Model Name

[curtischong]

Similar to how we can pass in a Fairchem model name:

model = FairChemModel(model="uma-s-1", task_name=task_name, cpu=DEVICE.type == "cpu")

and it will automatically find the pretrained model and automatically load it, we should do the same with MACE. I've always found having to find the proper imports to write:

# Load the MACE "small" foundation model
mace = mace_mp(model="small", return_raw_model=True)
mace_model = MaceModel(
    model=mace,
    device=device,
    dtype=torch.float64,
    compute_forces=True,
)

quite annoying. Making this change would also make our tutorials a bit shorter and easier to read!

[orionarcher]

Would be awesome to have!