Update README.md
Browse files
README.md
CHANGED
|
@@ -69,6 +69,26 @@ for more details, an re-download the correct version of the validation sets if n
|
|
| 69 |
| rattled-relax | 95,206 | [rattled-relax.tar.gz](https://dl.fbaipublicfiles.com/opencatalystproject/data/omat/241220/omat/val/rattled-relax.tar.gz) |
|
| 70 |
| Total | 1,025,361 | - |
|
| 71 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 72 |
### sAlex Dataset
|
| 73 |
We also provide the sAlex dataset used for fine-tuning of our OMat models. sAlex is a subsampled, Matbench-Discovery compliant, version of the original [Alexandria](https://alexandria.icams.rub.de/).
|
| 74 |
sAlex was created by removing structures matched in WBM and only sampling structure along a trajectory with an energy difference greater than 10 meV/atom. For full details,
|
|
|
|
| 69 |
| rattled-relax | 95,206 | [rattled-relax.tar.gz](https://dl.fbaipublicfiles.com/opencatalystproject/data/omat/241220/omat/val/rattled-relax.tar.gz) |
|
| 70 |
| Total | 1,025,361 | - |
|
| 71 |
|
| 72 |
+
##### <ins> 1M Subsplit </ins>
|
| 73 |
+
We provide a 1M training randomly subsampled split, and the corresponding validation and test so that it is easier to iterate and develop before training on the full dataset.
|
| 74 |
+
|
| 75 |
+
Download: [omat24_1M](https://dl.fbaipublicfiles.com/opencatalystproject/data/omat/251210/omat24_1M_251210.tar.gz)
|
| 76 |
+
|
| 77 |
+
| Train sub-dataset | Size |
|
| 78 |
+
| :-----: | :--: |
|
| 79 |
+
| rattled-1000 | 113,337 |
|
| 80 |
+
| rattled-1000-subsampled | 38,603 |
|
| 81 |
+
| rattled-500 | 68,885 |
|
| 82 |
+
| rattled-500-subsampled | 39,548 |
|
| 83 |
+
| rattled-300 | 62,894 |
|
| 84 |
+
| rattled-300-subsampled | 34,462 |
|
| 85 |
+
| aimd-from-PBE-1000-npt | 211,578 |
|
| 86 |
+
| aimd-from-PBE-1000-nvt | 201,512 |
|
| 87 |
+
| aimd-from-PBE-3000-npt | 60,477 |
|
| 88 |
+
| aimd-from-PBE-3000-nvt | 78,011 |
|
| 89 |
+
| rattled-relax | 100,543 |
|
| 90 |
+
| Total | 1,009,850 |
|
| 91 |
+
|
| 92 |
### sAlex Dataset
|
| 93 |
We also provide the sAlex dataset used for fine-tuning of our OMat models. sAlex is a subsampled, Matbench-Discovery compliant, version of the original [Alexandria](https://alexandria.icams.rub.de/).
|
| 94 |
sAlex was created by removing structures matched in WBM and only sampling structure along a trajectory with an energy difference greater than 10 meV/atom. For full details,
|