metadata
license: apache-2.0
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
- name: weight
dtype: float64
splits:
- name: train
num_bytes: 36626491
num_examples: 20000
download_size: 19631352
dataset_size: 36626491
configs:
- config_name: default
data_files:
- split: train
path: data/train-*