metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
- name: problem
dtype: string
- name: answer
dtype: string
- name: reasoning
dtype: string
- name: source
dtype: string
- name: reasoning_length
dtype: int64
- name: reasoning_length_log2
dtype: int64
splits:
- name: train
num_bytes: 94098648
num_examples: 5314
download_size: 52249234
dataset_size: 94098648
configs:
- config_name: default
data_files:
- split: train
path: data/train-*