SWCNT Dataset and CNN Models

Here are the dataset and machine learning models used for DNN training in the research paper:
Košmerl, V., Štajduhar, I., Čanađija, M.: "Predicting stress-strain behavior of carbon nanotubes using neural networks"
It contains basic data about 818 configurations and corresponding stress-strain values for all single-walled carbon nanotubes with diameters up to 4 nm, as well as CNN models for predicting stress-strain curves of SWCNTs. The dataset was obtained by averaging three sets of MD simulations in LAMMPS using modified AIREBO potential.
The column names are: n - chiral index, m - chiral index, D - actual diameter, L - actual length, D0 - theoretical diameter, L0 - initial length, DL - elongation. Other columns names are self-explanatory.
The TF-model predicts thermal fluctuations, while the S-model smooths them out. For all other details, please consult the above paper.
If you find this dataset and models useful, please cite the above paper. Full bibliographic data can be found at the listed DOI link.