xinyulu/vibench

introduction

This dataset contains the datasets used in our paper, Vib2Mol: from vibrational spectra to molecular structures—a versatile deep learning model. The full paper provides more details and is available on arXiv.
All datasets are pre-split for training, validation, and testing. To reproduce our results, you can find detailed steps, codes, and training logs in our GitHub repository.
A copy of the full dataset is also available for download on Figshare.

Demonstration

For demonstration purposes, this dataset card provides a copy of the VB-mols test set. Please note that this is a small demo set, not the entire dataset used for our research.

How to Use

After downloading the datasets, you can use the following Python code to visualize the data.

import lmdb
import pickle
import pandas as pd
from tqdm import tqdm

# Open the database
db = lmdb.open('path/to/lmdb/data', subdir=False, lock=False, map_size=int(1e11))

# Load data and convert to DataFrame
with db.begin() as txn:
    data = list(txn.cursor())

df = pd.DataFrame([pickle.loads(item[1]) for item in tqdm(data)])

# Now you can work with the DataFrame `df`
print(df.head())

Acknowledgements

This work was supported by the National Natural Science Foundation (Grant No: 22227802, 22021001, 22474117 and 22272139) of China and the Fundamental Research Funds for the Central Universities (20720220009 and 20720250005) and Shanghai Innovation Institute.

Contact

Welcome to contact us or raise issues if you have any questions. Email: xinyulu@stu.xmu.edu.cn