metadata

license: apache-2.0
size_categories:
  - 10K<n<100K
task_categories:
  - conversational
dataset_info:
  features:
    - name: text
      dtype: string
    - name: lang
      dtype: string
  splits:
    - name: train
      num_bytes: 23211220
      num_examples: 12947
  download_size: 13220375
  dataset_size: 23211220
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

OpenAssistant TOP-1 Conversation Threads

Guanacco style export of the best conversation threads from the open-assistant.io database
exported August 25, 2023
jsonl files with chatml formatted conversations
train: 12,947 samples
With a column indicating the language used.