Published March 26, 2026
| Version v2
Dataset
Open
Datasets for "Pushing the limits of one-dimensional NMR spectroscopy for automated structure elucidation using artificial intelligence"
Authors/Creators
-
1.
Stanford University
Description
This upload contains the datasets and checkpoints used for the experiments in:
Pushing the limits of one-dimensional NMR spectroscopy for automated structure elucidation using artificial intelligence
Frank Hu, Jonathan M. Tubb, Dimitris Argyropoulos, Sergey Golotvin, Mikhail Elyashberg, Grant M. Rotskoff, Matthew W. Kanan, and Thomas E. Markland
The zip files uploaded are as follows:
- experimental.zip: Training data and splits for BMRB experimental spectra and molecules
- pretraining.zip: Training data and splits for substructure-to-structure transformer training
- multitask.zip: Training data and splits for spectrum-to-structure/substructure multitask model training
- checkpoints.zip: Model checkpoints for all models used in the paper
For more descriptions of how to use each dataset, unzip the file and see the README contained.
Files
experimental.zip
Additional details
Software
- Repository URL
- https://github.com/MarklandGroup/NMRElucidator
- Programming language
- Python
- Development Status
- Active