---
dataset_info:
  features:
  - name: haiku
    dtype: string
  - name: source
    dtype: string
  splits:
  - name: train
    num_bytes: 11725740
    num_examples: 144123
  download_size: 7554208
  dataset_size: 11725740
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
license: cc-by-4.0
task_categories:
- text-generation
language:
- en
pretty_name: Haiku Dataset
---

source: https://www.kaggle.com/datasets/hjhalani30/haiku-dataset