Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
TorchANI 2.0 is an open-source library that supports training, development, and research of ANI-style neural network interatomic potentials. It was originally developed and is currently maintained by the Roitberg group.
| Date | Stars |
|---|---|
| 2026-07-31 | 553 |
| 2026-08-01 | 554 |
| 2026-08-05 | 554 |
| 2026-08-06 | 554 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
35.0
growth rate 0.00%/day
<div align="center">
<picture>
<source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/aiqm/torchani/main/front-logo-dark.png">
<source media="(prefers-color-scheme: light)" srcset="https://raw.githubusercontent.com/aiqm/torchani/main/front-logo-light.png">
<img alt="TorchANI 2 logo" src="https://raw.githubusercontent.com/aiqm/torchani/main/torchani-logo-light.png">
</picture>
</div>
[](https://opensource.org/licenses/MIT)
[](https://github.com/aiqm/torchani/actions/workflows/ci.yaml)
[](https://github.com/psf/black)
TorchANI 2.0 is an open-source library that supports training, development, and research
of ANI-style neural network interatomic potentials. It was originally developed and is
currently maintained by the Roitberg group. For information and examples, please see the
comprehensive [documentation](https://aiqm.github.io/torchani/).
⚠️ **Important**: If you were using a previous version of TorchANI and your code does not work with
TorchANI 2.0 check out the [migration guide](https://aiqm.github.io/torchani/migrating-to-2.html#torchani-migrating), there
are very few breaking changes, most code should work with minimal modifications. If
you can't figure something out please open a GitHub issue, we are here to help!
In the meantime, you can pin torchani to version 2.2.4 (pip install 'torchani==2.2.4'), which does not
have breaking changes. If you require the old state dicts of ANI models you can access
them by calling `.legacy_state_dict()` instead of `.state_dict()`
If you find a bug in TorchANI 2.0, or have some feature request, also feel free to open
a GitHub issue. TorchANI 2.0 is currently tested against PyTorch 2.13 and CUDA 13.0
If you find this work useful please cite the following articles:
- *TorchANI 2.0: An extensible, high performance library for the design, training, and use of NN-IPs* <br>
https://pubs.acs.org/doi/10.1021/acs.jcim.5c01853
- *TorchANI: A Free and Open Source PyTorch-Based Deep Learning Implementation of the ANI Neural Network Potentials* <br>
https://pubs.acs.org/doi/10.1021/acs.jcim.0c00451
To run molecular dynamics (full ML or ML/MM) with [Amber (sander or
pmemd)](https://ambermd.org/AmberTools.php) check out
[the TorchANI-Amber interface](https://github.com/roitberg-group/torchani-amber),
and the relevant publications:
- *TorchANI-Amber: Bridging neural network potentials and classical biomolecular simulations* <br>
https://doi.org/10.1021/acs.jpcb.5c05725
- *Advancing Multiscale Molecular Modeling with Machine Learning-Derived Electrostatics* <br>
For the ML/MM capabilities: https://pubs.acs.org/doi/10.1021/acs.jctc.4c01792
## Installation
We recommend installing `torchani` inside a `conda|mamba` environment, or a `venv`.
⚠️ **Important**: *Please install torchani with pip if you want the latest version, even
if using a conda env since the torchani conda package is currently not maintained.*
We also recommended you first install a specific torch version, with a specific CUDA
toolkit backend, for example:
```bash
pip install torch==2.13 --index-url https://download.pytorch.org/whl/cu130
```
for the version with CUDA 13.0. This is not strictly required, but is easier if you want
to control these versions. Note that TorchANI 2 requires PyTorch >= 2.0.
Afterwards:
```bash
pip install torchani
```
TorchANI 2.0 provides C++ and CUDA extensions for accelerated computation of descriptors
and network inference. In order to build the extensions, first install the CUDA Toolkit
appropriate for your PyTorch version. You can follow the instructions in [the official
documentation](https:Excerpt of 7,664 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:aa9e9a863259db7e, topic:deep-learning, topic:neural-network, topic:pytorch
