metadata

dataset_info:
  features:
    - name: conversation
      list:
        - name: content
          dtype: string
        - name: role
          dtype: string
  splits:
    - name: train
      num_bytes: 10491824
      num_examples: 5419
  download_size: 5658552
  dataset_size: 10491824
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

Dataset Card for "oasst2_top1_en"

More Information needed

Top 1% conversations of https://huggingface.co/datasets/OpenAssistant/oasst2
language-filtered: en
generated using https://github.com/blancsw/deep_4_all/blob/main/datasets/oasst/convert.py