metadata
dataset_info:
features:
- name: inst
dtype: string
- name: output
dtype: string
- name: id
dtype: string
- name: instruction
dtype: string
- name: input
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 735304357
num_examples: 50000
download_size: 427659214
dataset_size: 735304357
configs:
- config_name: default
data_files:
- split: train
path: data/train-*