metadata

dataset_info:
  features:
    - name: instruction
      dtype: string
    - name: response
      dtype: string
  splits:
    - name: train
      num_bytes: 310051
      num_examples: 423
  download_size: 176740
  dataset_size: 310051
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
tags:
  - synthetic
  - distilabel

Wikipedia ES Genstruct

This dataset contains a very small subset of instruction and generation pairs generated using Genstruct 7B AWQ and using some articles in Spanish from wikimedia/wikipedia as the raw corpus.