ChemQA / README.md
shangzhu's picture
Upload dataset
e1e0fff verified
metadata
language:
  - en
dataset_info:
  features:
    - name: image
      dtype: image
    - name: question
      dtype: string
    - name: choices
      dtype: string
    - name: label
      dtype: int64
    - name: description
      dtype: string
    - name: id
      dtype: string
  splits:
    - name: train
      num_bytes: 705885259.25
      num_examples: 66166
    - name: valid
      num_bytes: 100589192.25
      num_examples: 9486
    - name: test
      num_bytes: 100021131
      num_examples: 9480
  download_size: 866619578
  dataset_size: 906495582.5
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: valid
        path: data/valid-*
      - split: test
        path: data/test-*

Dataset Card for ChemQA

Introducing ChemQA: a Multimodal Question-and-Answering Dataset on Chemistry Reasoning. This work is inspired by IsoBench[1] and ChemLLMBench[2].

Content

There are 5 QA Tasks in total:

  • Counting Numbers of Carbons and Hydrogens in Organic Molecules: adapted from the 600 PubChem molecules created from [2], evenly divided into validation and evaluation datasets.
  • Calculating Molecular Weights in Organic Molecules: adapted from the 600 PubChem molecules created from [2], evenly divided into validation and evaluation datasets.
  • Name Conversion: From SMILES to IUPAC: adapted from the 600 PubChem molecules created from [2], evenly divided into validation and evaluation datasets.
  • Molecule Captioning and Editing: inspired by [2], adapted from dataset provided in [3], following the same training, validation and evaluation splits.
  • Retro-synthesis Planning: inspired by [2], adapted from dataset provided in [4], following the same training, validation and evaluation splits.

Load the Dataset

from datasets import load_dataset
dataset_train = load_dataset('shangzhu/ChemQA', split='train')
dataset_valid = load_dataset('shangzhu/ChemQA', split='valid')
dataset_test = load_dataset('shangzhu/ChemQA', split='test')

Reference

[1] Fu, D., Khalighinejad, G., Liu, O., Dhingra, B., Yogatama, D., Jia, R., & Neiswanger, W. (2024). IsoBench: Benchmarking Multimodal Foundation Models on Isomorphic Representations.

[2] Guo, T., Guo, kehan, Nan, B., Liang, Z., Guo, Z., Chawla, N., Wiest, O., & Zhang, X. (2023). What can Large Language Models do in chemistry? A comprehensive benchmark on eight tasks. Advances in Neural Information Processing Systems (Vol. 36, pp. 59662–59688).

[3] Edwards, C., Lai, T., Ros, K., Honke, G., Cho, K., & Ji, H. (2022). Translation between Molecules and Natural Language. Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, 375–413.

[4] Irwin, R., Dimitriadis, S., He, J., & Bjerrum, E. J. (2022). Chemformer: a pre-trained transformer for computational chemistry. Machine Learning: Science and Technology, 3(1), 15022.

Citation

@misc{chemQA2024,
      title={ChemQA: a Multimodal Question-and-Answering Dataset on Chemistry Reasoning}, 
      author={Shang Zhu and Xuefeng Liu and Ghazal Khalighinejad},
      year={2024},
      publisher={Hugging Face},
      howpublished={\url{https://huggingface.co/datasets/shangzhu/ChemQA}},
}

Contact

shangzhu@umich.edu