---
dataset_info:
  features:
  - name: drug_name
    dtype: string
  - name: application_number
    dtype: string
  - name: established_name
    dtype: string
  - name: chunk_index
    dtype: int64
  - name: section_heading
    dtype: string
  - name: page
    dtype: 'null'
  - name: prompt
    dtype: string
  - name: text
    dtype: string
  - name: extraction
    dtype: string
  - name: extraction_parsed_ok
    dtype: bool
  - name: reasoning
    dtype: string
  splits:
  - name: no_think
    num_bytes: 323978
    num_examples: 55
  - name: think
    num_bytes: 1712780
    num_examples: 55
  - name: thinking_v2
    num_bytes: 2466443
    num_examples: 75
  - name: thinking_v3
    num_bytes: 2524612
    num_examples: 75
  - name: thinking_v4
    num_bytes: 1671123
    num_examples: 53
  - name: opus_4_8
    num_bytes: 398339
    num_examples: 53
  - name: opus_4_8_v2
    num_bytes: 391580
    num_examples: 53
  - name: opus_4_8_v3
    num_bytes: 404689
    num_examples: 58
  - name: opus_4_8_v4
    num_bytes: 517848
    num_examples: 80
  - name: opus_4_8_back5000
    num_bytes: 53140143
    num_examples: 10121
  - name: opus_4_8_full
    num_bytes: 61083574
    num_examples: 10121
  download_size: 33077946
  dataset_size: 124635109
configs:
- config_name: default
  data_files:
  - split: no_think
    path: data/no_think-*
  - split: think
    path: data/think-*
  - split: thinking_v2
    path: data/thinking_v2-*
  - split: thinking_v3
    path: data/thinking_v3-*
  - split: thinking_v4
    path: data/thinking_v4-*
  - split: opus_4_8
    path: data/opus_4_8-*
  - split: opus_4_8_v2
    path: data/opus_4_8_v2-*
  - split: opus_4_8_v3
    path: data/opus_4_8_v3-*
  - split: opus_4_8_v4
    path: data/opus_4_8_v4-*
  - split: opus_4_8_full
    path: data/opus_4_8_full-*
  - split: opus_4_8_back5000
    path: data/opus_4_8_back5000-*
---