metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
- name: id
dtype: string
splits:
- name: train
num_bytes: 45419006
num_examples: 583
download_size: 18287282
dataset_size: 45419006
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
Original data from ClubFloyd, scraped by ve-forbryderne and hosted on W&B, I just took them and made them instruct-y