metadata
dataset_info:
features:
- name: conversations
list:
- name: content
dtype: string
- name: role
dtype: string
- name: metadata
struct:
- name: controversiality
dtype: int64
- name: normalized_controversiality
dtype: float64
- name: post
struct:
- name: author
dtype: string
- name: downvotes
dtype: int64
- name: flair
dtype: string
- name: score
dtype: int64
- name: suggested_sort
dtype: string
- name: upvote_ratio
dtype: float64
- name: upvotes
dtype: int64
- name: subreddit
struct:
- name: name
dtype: string
- name: subscribers
dtype: int64
splits:
- name: train
num_bytes: 279326.56871928705
num_examples: 218
download_size: 132052
dataset_size: 279326.56871928705
configs:
- config_name: default
data_files:
- split: train
path: data/train-*