README.md

metadata

language:
  - en
size_categories:
  - 1M<n<10M
task_categories:
  - text-classification
pretty_name: CoconutDB SMILES to Formula Pair Classification
dataset_info:
  features:
    - name: formula
      dtype: string
    - name: smiles
      dtype: string
    - name: label
      dtype: int64
  splits:
    - name: test
      num_bytes: 457537
      num_examples: 4000
  download_size: 210648
  dataset_size: 457537
configs:
  - config_name: default
    data_files:
      - split: test
        path: data/test-*
tags:
  - chemistry
  - coconutdb
  - SMILES
  - chemteb

CoconutDB SMILES to Formula Pair Classification

This dataset contains pairs of SMILES strings (both isomeric and canonical) and their corresponding molecular formulas, with labels indicating whether they refer to the same chemical entity. A label of 1 means the SMILES string and the molecular formula correspond to the same entity, while a label of 0 indicates they do not. The dataset is sourced from CoconutDB and provides valuable information for tasks involving chemical entity matching and molecular formula analysis.