metadata
dataset_info:
features:
- name: text
dtype: string
- name: trace
sequence: string
- name: steps
dtype: string
- name: len_target
dtype: int64
- name: query
dtype: string
- name: target
dtype: string
- name: paths
sequence: string
splits:
- name: train
num_bytes: 85778052
num_examples: 90000
- name: validation
num_bytes: 9553914
num_examples: 10000
download_size: 57024544
dataset_size: 95331966
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*