metadata
dataset_info:
features:
- name: Category
dtype: string
- name: input
dtype: string
- name: response
dtype: string
- name: conversations
dtype: string
- name: text
dtype: string
- name: source
dtype: string
- name: score
dtype: float64
- name: reasoning_process
dtype: string
splits:
- name: train
num_bytes: 123578565
num_examples: 50000
download_size: 1737405
dataset_size: 123578565
configs:
- config_name: default
data_files:
- split: train
path: data/train-*