---
dataset_info:
- config_name: Original
  features:
  - name: DataBase
    list:
    - name: id
      dtype: string
    - name: sql
      dtype: binary
    - name: sqlite
      dtype: binary
  - name: Dataset
    struct:
    - name: identifier
      dtype: string
    - name: name
      dtype: string
    - name: src_path
      dtype: string
    - name: label_path
      dtype: string
    - name: category
      dtype: int32
    - name: type
      dtype: int32
  - name: data
    list:
    - name: db_id
      dtype: string
    - name: utterance_id
      dtype: string
    - name: hardness
      dtype: string
    - name: utterance_type
      dtype: string
    - name: query
      dtype: string
    - name: utterance
      dtype: string
    - name: values
      list:
      - name: token
        dtype: string
      - name: start
        dtype: int32
      - name: column_index
        dtype: int32
    - name: cols
      list:
      - name: token
        dtype: string
      - name: start
        dtype: int32
      - name: column_index
        dtype: int32
  splits:
  - name: train
    num_bytes: 30443052
    num_examples: 17
  - name: validation
    num_bytes: 3773264
    num_examples: 17
  download_size: 8501291
  dataset_size: 34216316
- config_name: SFT
  features:
  - name: id
    dtype: string
  - name: prompt
    dtype: string
  - name: answer
    dtype: string
  - name: metadata
    struct:
    - name: hardness
      dtype: string
    - name: values
      list:
      - name: token
        dtype: string
      - name: start
        dtype: int32
      - name: column_index
        dtype: int32
    - name: cols
      list:
      - name: token
        dtype: string
      - name: start
        dtype: int32
      - name: column_index
        dtype: int32
  splits:
  - name: train
    num_bytes: 29462552
    num_examples: 88946
  - name: validation
    num_bytes: 3649884
    num_examples: 11026
  download_size: 9392031
  dataset_size: 33112436
configs:
- config_name: Original
  data_files:
  - split: train
    path: Original/train-*
  - split: validation
    path: Original/validation-*
- config_name: SFT
  data_files:
  - split: train
    path: SFT/train-*
  - split: validation
    path: SFT/validation-*
---