---
dataset_info:
  features:
  - name: info
    struct:
    - name: article
      dtype: 'null'
    - name: id
      dtype: string
    - name: post
      dtype: string
    - name: site
      dtype: 'null'
    - name: subreddit
      dtype: string
    - name: title
      dtype: string
  - name: summaries
    list:
    - name: note
      dtype: string
    - name: policy
      dtype: string
    - name: text
      dtype: string
  - name: choice
    dtype: int64
  - name: worker
    dtype: string
  - name: batch
    dtype: string
  - name: split
    dtype: string
  - name: extra
    struct:
    - name: confidence
      dtype: int64
  - name: query_token
    sequence: int64
  - name: query
    dtype: string
  - name: chosen
    dtype: string
  - name: chosen_token
    sequence: int64
  - name: chosen_token_len
    dtype: int64
  - name: rejected
    dtype: string
  - name: rejected_token
    sequence: int64
  - name: rejected_token_len
    dtype: int64
  - name: chosen_policy
    dtype: string
  - name: rejected_policy
    dtype: string
  - name: policies
    dtype: string
  - name: query_chosen
    dtype: string
  - name: query_chosen_token
    sequence: int64
  - name: query_chosen_token_len
    dtype: int64
  - name: query_rejected
    dtype: string
  - name: query_rejected_token
    sequence: int64
  - name: query_rejected_token_len
    dtype: int64
  - name: query_token_len
    dtype: int64
  - name: query_chosen_token_response_label
    sequence: int64
  - name: query_rejected_token_response_label
    sequence: int64
  - name: chosen_score
    dtype: float64
  - name: rejected_score
    dtype: float64
  - name: swap
    dtype: bool
  splits:
  - name: train
    num_bytes: 3161418711
    num_examples: 92858
  - name: test
    num_bytes: 774120615
    num_examples: 22637
  download_size: 175892513
  dataset_size: 3935539326
---
# Dataset Card for "summarize_from_feedback_oai_preprocessing_pythia-6.9b-gold"

[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)