---
dataset_info:
  features:
  - name: original_query
    dtype: string
  - name: original_answer
    dtype: string
  - name: refinement_sequence
    list:
    - name: iteration
      dtype: int64
    - name: refined_query
      dtype: string
    - name: refinement_type
      dtype: string
    - name: retrieved_docs
      sequence: string
  - name: new_answer
    dtype: string
  - name: input_ids
    sequence: int32
  - name: attention_mask
    sequence: int8
  splits:
  - name: train
    num_bytes: 289729
    num_examples: 100
  download_size: 11798
  dataset_size: 289729
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---