---
dataset_info:
  config_name: multitask
  features:
  - name: text
    dtype: string
  - name: choices
    sequence: string
  - name: label
    dtype: int64
  splits:
  - name: train
    num_bytes: 480162837
    num_examples: 1053588
  download_size: 262013771
  dataset_size: 480162837
configs:
- config_name: multitask
  data_files:
  - split: train
    path: multitask/train-*
license: odc-by
language:
- en
size_categories:
- 1M<n<10M
task_categories:
- text2text-generation
---

# taskweb

- https://github.com/danieljkim0118/TaskWeb
- https://arxiv.org/abs/2305.13256