---
dataset_info:
  features:
  - name: id
    dtype: string
  - name: profile
    struct:
    - name: age
      struct:
      - name: day
        dtype: int64
      - name: month
        dtype: int64
      - name: week
        dtype: int64
      - name: year
        dtype: int64
    - name: gender
      dtype: string
    - name: height
      dtype: int64
    - name: medical_history
      dtype: string
    - name: occupation
      dtype: string
    - name: oldcarts
      struct:
      - name: aggravating
        dtype: string
      - name: characteristic
        dtype: string
      - name: duration
        dtype: string
      - name: location
        dtype: string
      - name: onset
        dtype: string
      - name: reliving_factors
        dtype: string
      - name: severity
        dtype: string
      - name: timing
        dtype: string
    - name: presenting_complaint
      dtype: string
    - name: primary_symptom
      dtype: string
    - name: secondary_symptoms
      list: string
    - name: underlying_diseases
      list: string
    - name: weight
      dtype: float64
  - name: label
    struct:
    - name: department
      struct:
      - name: description
        dtype: string
      - name: id
        dtype: string
      - name: name
        dtype: string
      - name: name_th
        dtype: string
    - name: disease
      struct:
      - name: description
        dtype: string
      - name: full_name
        dtype: string
      - name: id
        dtype: string
      - name: name
        dtype: string
      - name: name_th
        dtype: string
    - name: severity
      struct:
      - name: description
        dtype: string
      - name: id
        dtype: string
      - name: name
        dtype: string
      - name: name_th
        dtype: string
  - name: usage
    struct:
    - name: approx_cost_usd
      dtype: float64
    - name: completion_tokens
      dtype: int64
    - name: model_name
      dtype: string
    - name: prompt_tokens
      dtype: int64
    - name: reasoning_tokens
      dtype: int64
  splits:
  - name: train
    num_bytes: 3830860
    num_examples: 1250
  download_size: 955376
  dataset_size: 3830860
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---