metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
- name: run_id
dtype: string
- name: step
dtype: int64
- name: uid
dtype: int64
- name: id
dtype: string
splits:
- name: train
num_bytes: 29254026.08151392
num_examples: 12656
download_size: 15218406
dataset_size: 29254026.08151392
configs:
- config_name: default
data_files:
- split: train
path: data/train-*