Update README.md
Browse files
README.md
CHANGED
|
@@ -1,10 +1,51 @@
|
|
| 1 |
-
|
| 2 |
-
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Isayev Lab at CMU
|
| 2 |
+
|
| 3 |
+
We develop machine learning methods for chemistry and materials science at [Carnegie Mellon University](https://www.cmu.edu/). Our research spans neural network potentials, molecular generation, and AI-driven discovery of molecules and materials.
|
| 4 |
+
|
| 5 |
+
**PI:** [Olexandr Isayev](https://scholar.google.com/citations?user=GnATMRoAAAAJ)
|
| 6 |
+
|
| 7 |
+
## AIMNet2 Models
|
| 8 |
+
|
| 9 |
+
AIMNet2 is our neural network interatomic potential for fast, accurate molecular simulations. It predicts energies, forces, charges, and Hessians at near-DFT accuracy, orders of magnitude faster.
|
| 10 |
+
|
| 11 |
+
| Model | Description | Elements |
|
| 12 |
+
|-------|-------------|----------|
|
| 13 |
+
| [aimnet2-wb97m-d3](https://huggingface.co/isayevlab/aimnet2-wb97m-d3) | General purpose (wB97M-D3) | 14: H,B,C,N,O,F,Si,P,S,Cl,As,Se,Br,I |
|
| 14 |
+
| [aimnet2-2025](https://huggingface.co/isayevlab/aimnet2-2025) | Improved intermolecular (B97-3c) | 14: H,B,C,N,O,F,Si,P,S,Cl,As,Se,Br,I |
|
| 15 |
+
| [aimnet2-nse](https://huggingface.co/isayevlab/aimnet2-nse) | Open-shell / radicals (wB97M) | 14: H,B,C,N,O,F,Si,P,S,Cl,As,Se,Br,I |
|
| 16 |
+
| [aimnet2-pd](https://huggingface.co/isayevlab/aimnet2-pd) | Palladium chemistry (wB97M-D3) | 14: H,B,C,N,O,F,Si,P,S,Cl,Se,Br,Pd,I |
|
| 17 |
+
|
| 18 |
+
All models include 4 ensemble members in safetensors format for uncertainty estimation.
|
| 19 |
+
|
| 20 |
+
## Quick Start
|
| 21 |
+
|
| 22 |
+
```bash
|
| 23 |
+
pip install "aimnet[hf]"
|
| 24 |
+
```
|
| 25 |
+
|
| 26 |
+
```python
|
| 27 |
+
from aimnet.calculators import AIMNet2Calculator
|
| 28 |
+
|
| 29 |
+
calc = AIMNet2Calculator("isayevlab/aimnet2-wb97m-d3")
|
| 30 |
+
results = calc(
|
| 31 |
+
{"coord": coords, "numbers": atomic_numbers, "charge": 0.0},
|
| 32 |
+
forces=True,
|
| 33 |
+
)
|
| 34 |
+
```
|
| 35 |
+
|
| 36 |
+
## Links
|
| 37 |
+
|
| 38 |
+
- [GitHub](https://github.com/isayevlab) | [Lab Website](https://isayevlab.org/) | [AIMNet2 Paper](https://doi.org/10.1039/D4SC08572H)
|
| 39 |
+
|
| 40 |
+
## Citation
|
| 41 |
+
|
| 42 |
+
```bibtex
|
| 43 |
+
@article{anstine2025aimnet2,
|
| 44 |
+
title={AIMNet2: A Neural Network Potential to Meet your Neutral, Charged, Organic, and Elemental-Organic Needs},
|
| 45 |
+
author={Anstine, Dylan and Zubatyuk, Roman and Isayev, Olexandr},
|
| 46 |
+
journal={Chemical Science},
|
| 47 |
+
year={2025},
|
| 48 |
+
publisher={Royal Society of Chemistry},
|
| 49 |
+
doi={10.1039/D4SC08572H}
|
| 50 |
+
}
|
| 51 |
+
```
|