Datasets:
Tasks:
Text2Text Generation
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
10K - 100K
Tags:
recipe
License:
metadata
license: afl-3.0
task_categories:
- text2text-generation
language:
- en
tags:
- recipe
size_categories:
- 10M<n<100M
source_datasets:
- lishuyang/recipepairs
configs:
- config_name: default
data_files:
- split: train
path: pair_data_train.jsonl
- split: test
path: pair_data_test.jsonl
- split: validation
path: pair_data_val.jsonl
train : 64K pairs test&validate : 8K pairs 8 : 1 : 1