Garden's Almanac of Matter Models | Checkpoint | Params | Polaris | Sophia | Perlmutter | Delta | Frontier | Della |
|---|---|---|---|---|---|---|---|
| tece-oam-rra-1.0 | 222M | β | β | β | β | β | β |
β verified, last 30 days β installed, not recently verified β not installed
No cluster has installed any checkpoint of this model yet.
Running this model
1# from a job or interactive session on a supported cluster:
2from rootstock import RootstockCalculator
3
4with RootstockCalculator(
5 cluster=YOUR_CLUSTER_ID, # eg, "delta"
6 checkpoint="tece-oam-rra-1.0",
7 device="cuda",
8) as calc:
9 # model is now running in subprocess on compute node
10 atoms.calc = calc
11 atoms.get_potential_energy()
Environments
Rootstock runs each model family inside an isolated Python environment defined by a single file. This environment file includes the specific dependencies needed, plus a setup() function that loads the model and returns an ASE calculator. These files are usually almost identical for a given model family, but because of cluster-specific quirks (eg, an old CUDA driver) the dependencies and setup code can vary a bit.
tace_env.py
1# /// script
2# requires-python = ">=3.11"
3# dependencies = [
4# "tace",
5# "ase>=3.22",
6# "torch>=2.4,<2.14",
7# ]
8#
9# # Delta's driver caps at CUDA 12.8 β route torch to the cu128 index, or the
10# # default PyPI wheel (cu130 for torch >=2.9) fails CUDA init on every checkpoint.
11# [tool.uv.sources]
12# torch = { index = "pytorch-cu128" }
13#
14# # The TECE architecture and its foundation registry ship in tace 0.2.0,
15# # which is not on PyPI (latest release there is 0.1.0). This is the commit
16# # the Matbench Discovery submission pinned.
17# [tool.uv.sources.tace]
18# git = "https://github.com/xvzemin/tace"
19# rev = "81f65a4c188bd09cec8d1419388f7afdcc1b6fd0"
20#
21# [[tool.uv.index]]
22# name = "pytorch-cu128"
23# url = "https://download.pytorch.org/whl/cu128"
24# explicit = true
25# ///
26"""TACE env β hosts TECE/TACE foundation checkpoints (Xu, Xie & Hu).
27
28openequivariance CUDA-kernel acceleration is optional upstream and not
29installed here β TACE runs on the e3nn path.
30"""
31
32CHECKPOINTS = {
33 "tece-oam-rra-1.0": "TECE-OAM-RRA-1.0",
34}
35
36
37def setup(checkpoint: str, device: str = "cuda"):
38 from tace.foundations import tace_foundations
39 from tace.interface.ase import TACEAseCalc
40
41 model_path = tace_foundations[CHECKPOINTS[checkpoint]]
42 return TACEAseCalc(model_path, device=device, dtype="float32")
43
Built on Delta: 2026-08-04
Couldn't load the current environments from Rootstock.