metadata

dataset_info:
  features:
    - name: q
      dtype: string
    - name: a
      dtype: string
  splits:
    - name: train
      num_bytes: 2441555
      num_examples: 3200
  download_size: 998771
  dataset_size: 2441555
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: cc-by-sa-4.0
task_categories:
  - text-generation
language:
  - th
size_categories:
  - 1K<n<10K

Dataset Card for Han Instruct Dataset v2.0

🪿 Han (ห่าน or goose) Instruct Dataset is a Thai instruction dataset by PyThaiNLP. This dataset collect all Thai instruct dataset that made by human and our old model. The dataset can use to train Instruction Following model like ChatGPT or other.

Many question are collect from Reference desk at Thai wikipedia.

Data sources:

Reference desk at Thai wikipedia.
Law from justicechannel.org
pythainlp/final_training_set_v1_enth: Human checked and edited.
Self-instruct from WangChanGLM
Wannaphong.com
Blognone
Human annotators

Supported Tasks and Leaderboards

ChatBot
Instruction Following

Languages

Thai

Dataset Structure

Data Fields

q: Question
a: Answer

Considerations for Using the Data

The dataset can has a biased from human annotators. We recommend you should checked the dataset to select or remove an instruct follow before train the model or use it own your risk.

Licensing Information

CC-BY-SA 4.0

Citation

If you use Han Instruct Dataset (v2.0) in your project or publication, please cite the dataset as follows:

Phatthiyaphaibun, W. (2024). Han Instruct Dataset (v2.0) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.10935857

@dataset{phatthiyaphaibun_2024_10935857,
  author       = {Phatthiyaphaibun, Wannaphong},
  title        = {Han Instruct Dataset},
  month        = apr,
  year         = 2024,
  publisher    = {Zenodo},
  version      = {v2.0},
  doi          = {10.5281/zenodo.10935857},
  url          = {https://doi.org/10.5281/zenodo.10935857}
}

Zenodo: https://doi.org/10.5281/zenodo.10935857