metadata
dataset_info:
features:
- name: text
dtype: string
- name: total_tokens
dtype: int64
- name: assistant_tokens
dtype: int64
- name: model
dtype: string
- name: assistant_replies
dtype: int64
- name: num_actors
dtype: int64
- name: total_replies
dtype: int64
- name: actors
sequence: string
- name: tools
sequence: string
splits:
- name: train
num_bytes: 62514
num_examples: 40
download_size: 46849
dataset_size: 62514
configs:
- config_name: default
data_files:
- split: train
path: data/train-*