metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
- name: weight
dtype: float64
- name: modified_text
dtype: string
- name: extracted_transcript
dtype: string
- name: original_text
dtype: string
- name: original_response
dtype: string
- name: system
dtype: string
- name: generated_id
dtype: string
splits:
- name: train
num_bytes: 28951931
num_examples: 9150
download_size: 16041787
dataset_size: 28951931
configs:
- config_name: default
data_files:
- split: train
path: data/train-*