---
dataset_info:
  features:
  - name: prompt_id
    dtype: string
  - name: anchor
    dtype: string
  - name: response
    dtype: string
  - name: cosine_similarity
    dtype: float64
  splits:
  - name: train
    num_bytes: 26382764
    num_examples: 4992
  download_size: 15643681
  dataset_size: 26382764
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---