README.md

---
dataset_info:
  features:
  - name: 'Unnamed: 0'
    dtype: int64
  - name: level_0
    dtype: int64
  - name: index
    dtype: float64
  - name: original_rxn
    dtype: string
  - name: fragments
    dtype: string
  - name: source
    dtype: string
  - name: year
    dtype: int64
  - name: mapped_rxn
    dtype: string
  - name: confidence
    dtype: float64
  - name: canonical_rxn_with_fragment_info
    dtype: string
  - name: canonical_rxn
    dtype: string
  - name: ID
    dtype: string
  - name: reaction_hash
    dtype: string
  - name: reactants
    dtype: string
  - name: products
    dtype: string
  - name: retro_template
    dtype: string
  - name: template_hash
    dtype: string
  - name: selectivity
    dtype: float64
  - name: outcomes
    dtype: float64
  - name: reagents
    dtype: string
  - name: labels
    dtype: int64
  splits:
  - name: train
    num_bytes: 754334119
    num_examples: 400604
  - name: test
    num_bytes: 83993127
    num_examples: 44511
  download_size: 363489916
  dataset_size: 838327246
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---