metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: chosen
list:
- name: content
dtype: string
- name: role
dtype: string
- name: rejected
list:
- name: content
dtype: string
- name: role
dtype: string
- name: index
dtype: int64
- name: medical_tier
dtype: string
- name: reasoning_for_medical_tier
dtype: string
- name: medical_classification_full_response
dtype: string
- name: medical_classification_model_id
dtype: string
splits:
- name: train
num_bytes: 7065304
num_examples: 3240
download_size: 3654851
dataset_size: 7065304
configs:
- config_name: default
data_files:
- split: train
path: data/train-*