Datasets:

pkanithi
/

ChEBI-20-MM-SMILES-Captioning

dataset_info:
  features:
    - name: CID
      dtype: int64
    - name: SMILES
      dtype: string
    - name: description
      dtype: string
    - name: polararea
      dtype: float64
    - name: xlogp
      dtype: float64
    - name: inchi
      dtype: string
    - name: iupacname
      dtype: string
    - name: SELFIES
      dtype: string
    - name: messages
      list:
        - name: content
          dtype: string
        - name: role
          dtype: string
  splits:
    - name: train
      num_bytes: 42109399
      num_examples: 26402
    - name: validation
      num_bytes: 5291676
      num_examples: 3299
    - name: test
      num_bytes: 5200785
      num_examples: 3297
  download_size: 19486426
  dataset_size: 52601860
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*

ChEBI-20-MM SMILES Captioning Dataset

Dataset Transformation Summary

Original Dataset: liupf/ChEBI-20-MM
Transformed Dataset: pkanithi/ChEBI-20-MM-SMILES-Captioning

Changes Made

Added `messages` Column

Added a new messages column in ChatML format
Each example now contains a conversation structure with system, user, and assistant messages

Chat Format Structure

Each example now has a messages array with:

System message: Instructions for molecular structure description
User message: SMILES string input
Assistant message: Molecular description from the dataset

The user message format:

Here is the molecular data:
- SMILES: [SMILES_string]

Result

The dataset is now in ChatML format suitable for supervised fine-tuning of molecular captioning models, focusing on SMILES-to-description tasks.