metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: completion
dtype: string
splits:
- name: train
num_bytes: 91035395
num_examples: 142622
download_size: 42482187
dataset_size: 91035395
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
Dataset Card for "VMware-open-instruct-higgsfield"
Just a port of the VMware/open-instruct dataset to Higgsfield's prompt-completion format.