metadata
dataset_info:
features:
- name: source
dtype: string
- name: target
dtype: string
- name: target_backto_source
dtype: string
- name: raw_target
list:
- name: generated_text
dtype: string
- name: raw_target_backto_source
list:
- name: generated_text
dtype: string
- name: prompt
dtype: string
- name: reverse_prompt
dtype: string
- name: source_langid
dtype: string
- name: target_langid
dtype: string
- name: target_backto_source_langid
dtype: string
- name: doc_id
dtype: int64
- name: sent_id
dtype: int64
- name: timestamp
dtype: string
- name: url
dtype: string
- name: doc_hash
dtype: string
splits:
- name: train
num_bytes: 3885
num_examples: 4
download_size: 14380
dataset_size: 3885
configs:
- config_name: default
data_files:
- split: train
path: b88d87373fe12f07/train-*