---
dataset_info:
  features:
  - name: prompt
    dtype: string
  - name: responses
    list: string
  - name: train
    dtype: string
  - name: test
    dtype: string
  - name: source
    dtype: string
  - name: answer
    dtype: string
  - name: num_tokens
    dtype: int64
  - name: avg_hamming_distance
    dtype: float64
  - name: avg_percentage_match
    dtype: float64
  - name: avg_is_correct
    dtype: float64
  - name: avg_response_length
    dtype: float64
  - name: hamming_distances
    list: int64
  - name: percentage_matches
    list: float64
  - name: is_corrects
    list: bool
  - name: response_lengths
    list: int64
  - name: extracted_responses
    list:
      list:
        list: int64
  splits:
  - name: train
    num_bytes: 4201029629
    num_examples: 11169
  - name: test
    num_bytes: 49394048
    num_examples: 120
  download_size: 1472928923
  dataset_size: 4250423677
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---