---
dataset_info:
  features:
  - name: question_id
    dtype: string
  - name: sample_id
    dtype: int64
  - name: prompt_pretokenized
    dtype: string
  - name: processed_response
    dtype: string
  - name: unprocessed_response
    dtype: string
  - name: slightly_processed_response
    dtype: string
  - name: accuracy
    dtype: float64
  - name: rm_score
    dtype: float64
  splits:
  - name: train
    num_bytes: 5939310198
    num_examples: 1508687
  download_size: 1000852072
  dataset_size: 5939310198
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---