metadata
dataset_info:
features:
- name: id
dtype: string
- name: question
dtype: string
- name: answer
dtype: string
- name: topic
dtype: string
- name: parent_seed_ids
dtype: string
- name: target_molecule
dtype: string
- name: candidate_molecule
dtype: string
- name: cohort_vector
dtype: string
- name: cohort_target
dtype: string
- name: merge_complexity
dtype: string
- name: audit_json
dtype: string
- name: clean_accept
dtype: int64
- name: refinement_rounds
dtype: int64
- name: accepted_at
dtype: float64
splits:
- name: train
num_bytes: 36500
num_examples: 10
download_size: 31652
dataset_size: 36500
configs:
- config_name: default
data_files:
- split: train
path: data/train-*