# CGCNN2
[](https://github.com/jcwang587/cgcnn2/releases/latest)
[](https://pepy.tech/projects/cgcnn2)
<!--
[](https://github.com/astral-sh/ruff)
[](https://results.pre-commit.ci/latest/github/jcwang587/cgcnn2/main)
-->
As the original Crystal Graph Convolutional Neural Networks (CGCNN) repository is no longer actively maintained, this repository is a reproduction of [CGCNN](https://github.com/txie-93/cgcnn) by Xie *et al*. It includes necessary updates for deprecated components and a few additional functions to ensure smooth operation. Despite its age, CGCNN remains a straightforward and fast deep learning framework that is easy to learn and use.
The package provides the following major functions:
- **Training** a CGCNN model using a custom dataset.
- **Predicting** material properties with a pre-trained CGCNN model.
- **Fine-tuning** a pre-trained CGCNN model on a new dataset.
- **Extracting** structural features as descriptors for downstream tasks.
<!---**Augmenting** training data by pertubing atomic positions (in development).-->
## Installation
Make sure you have a Python interpreter, preferably version 3.11 or higher. Then, you can simply install cgcnn2 from
PyPI using `pip`:
```bash
pip install cgcnn2
```
If you'd like to use the latest unreleased version on the main branch, you can install it directly from GitHub:
```bash
pip install git+https://github.com/jcwang587/cgcnn2@main
```
## Get Started
There are entry points for training, predicting, and fine-tuning CGCNN models. For example, to explore the usage of the provided training script `cgcnn-tr`, you can use the `--help` option of the command:
```bash
cgcnn-tr --help
```
Similarly, you can access the predicting and fine-tuning functionalities through `cgcnn-pr` and `cgcnn-ft` commands. A detailed user guide documentation is available at: [https://jcwang.dev/cgcnn2/](https://jcwang.dev/cgcnn2/)
## References
The original paper describes the CGCNN framework in detail:
```bibtex
@article{PhysRevLett2018,
title = {Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties},
author = {Xie, Tian and Grossman, Jeffrey C.},
journal = {Phys. Rev. Lett.},
volume = {120},
issue = {14},
pages = {145301},
numpages = {6},
year = {2018},
month = {Apr},
publisher = {American Physical Society},
doi = {10.1103/PhysRevLett.120.145301},
url = {https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.120.145301}
}
```
Raw data
{
"_id": null,
"home_page": null,
"name": "cgcnn2",
"maintainer": "Jiacheng Wang",
"docs_url": null,
"requires_python": "<3.13,>=3.11",
"maintainer_email": "Jiacheng Wang <jiachengwang@umass.edu>",
"keywords": "computational-materials, fine-tuning, graph-model, python",
"author": "Jiacheng Wang",
"author_email": "Jiacheng Wang <jiachengwang@umass.edu>",
"download_url": "https://files.pythonhosted.org/packages/00/f7/9635742b4d7606db9b43b8ab35ba219f28c7fccbf070c8423e418fdf7548/cgcnn2-0.5.2.tar.gz",
"platform": null,
"description": "# CGCNN2\n\n[](https://github.com/jcwang587/cgcnn2/releases/latest)\n[](https://pepy.tech/projects/cgcnn2)\n\n<!--\n[](https://github.com/astral-sh/ruff)\n\n[](https://results.pre-commit.ci/latest/github/jcwang587/cgcnn2/main)\n-->\n\nAs the original Crystal Graph Convolutional Neural Networks (CGCNN) repository is no longer actively maintained, this repository is a reproduction of [CGCNN](https://github.com/txie-93/cgcnn) by Xie *et al*. It includes necessary updates for deprecated components and a few additional functions to ensure smooth operation. Despite its age, CGCNN remains a straightforward and fast deep learning framework that is easy to learn and use.\n\nThe package provides the following major functions:\n\n- **Training** a CGCNN model using a custom dataset.\n- **Predicting** material properties with a pre-trained CGCNN model.\n- **Fine-tuning** a pre-trained CGCNN model on a new dataset.\n- **Extracting** structural features as descriptors for downstream tasks.\n\n<!---**Augmenting** training data by pertubing atomic positions (in development).-->\n\n## Installation\n\nMake sure you have a Python interpreter, preferably version 3.11 or higher. Then, you can simply install cgcnn2 from\nPyPI using `pip`:\n\n```bash\npip install cgcnn2\n```\n\nIf you'd like to use the latest unreleased version on the main branch, you can install it directly from GitHub:\n\n```bash\npip install git+https://github.com/jcwang587/cgcnn2@main\n```\n\n## Get Started\n\nThere are entry points for training, predicting, and fine-tuning CGCNN models. For example, to explore the usage of the provided training script `cgcnn-tr`, you can use the `--help` option of the command:\n\n```bash\ncgcnn-tr --help\n```\n\nSimilarly, you can access the predicting and fine-tuning functionalities through `cgcnn-pr` and `cgcnn-ft` commands. A detailed user guide documentation is available at: [https://jcwang.dev/cgcnn2/](https://jcwang.dev/cgcnn2/)\n\n## References\n\nThe original paper describes the CGCNN framework in detail:\n\n```bibtex\n@article{PhysRevLett2018,\n title = {Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties},\n author = {Xie, Tian and Grossman, Jeffrey C.},\n journal = {Phys. Rev. Lett.},\n volume = {120},\n issue = {14},\n pages = {145301},\n numpages = {6},\n year = {2018},\n month = {Apr},\n publisher = {American Physical Society},\n doi = {10.1103/PhysRevLett.120.145301},\n url = {https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.120.145301}\n}\n```\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "Reproduction of Crystal Graph Convolutional Neural Networks",
"version": "0.5.2",
"project_urls": null,
"split_keywords": [
"computational-materials",
" fine-tuning",
" graph-model",
" python"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "d30467c0e1e4f026366fa5d9ddc56665c18e77873bcf9776681e8051434546c9",
"md5": "34ae2c7f59ab3a1505bae31c92f6c6d9",
"sha256": "96f4da04ae17c0e5930a68f6782580d922ad8163843bd2b9821662d8b7148609"
},
"downloads": -1,
"filename": "cgcnn2-0.5.2-py3-none-any.whl",
"has_sig": false,
"md5_digest": "34ae2c7f59ab3a1505bae31c92f6c6d9",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": "<3.13,>=3.11",
"size": 5879790,
"upload_time": "2025-08-13T20:19:03",
"upload_time_iso_8601": "2025-08-13T20:19:03.694018Z",
"url": "https://files.pythonhosted.org/packages/d3/04/67c0e1e4f026366fa5d9ddc56665c18e77873bcf9776681e8051434546c9/cgcnn2-0.5.2-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "00f79635742b4d7606db9b43b8ab35ba219f28c7fccbf070c8423e418fdf7548",
"md5": "c5926cd370ae3f62f7dc6792d25e4125",
"sha256": "79ea9156a190c75e6ab560713eb37a6663dd2a1da73e34401ceecb000257e78e"
},
"downloads": -1,
"filename": "cgcnn2-0.5.2.tar.gz",
"has_sig": false,
"md5_digest": "c5926cd370ae3f62f7dc6792d25e4125",
"packagetype": "sdist",
"python_version": "source",
"requires_python": "<3.13,>=3.11",
"size": 5874363,
"upload_time": "2025-08-13T20:19:07",
"upload_time_iso_8601": "2025-08-13T20:19:07.027876Z",
"url": "https://files.pythonhosted.org/packages/00/f7/9635742b4d7606db9b43b8ab35ba219f28c7fccbf070c8423e418fdf7548/cgcnn2-0.5.2.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-08-13 20:19:07",
"github": false,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"lcname": "cgcnn2"
}
JSON