---
dataset_info:
  features:
  - name: source
    dtype: string
  - name: source_dataset
    dtype: string
  - name: quality_metrics
    dtype: 'null'
  - name: prompt
    dtype: string
  - name: response
    dtype: string
  - name: answer
    dtype: string
  - name: int_answer
    dtype: string
  - name: input_ids
    list: int32
  - name: attention_mask
    list: int8
  - name: bucket_id
    dtype: int64
  - name: answer_token
    dtype: int64
  splits:
  - name: train
    num_bytes: 1752779673
    num_examples: 670282
  download_size: 767050621
  dataset_size: 1752779673
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---