metadata
dataset_info:
features:
- name: id
dtype: int64
- name: text
dtype: string
- name: reasoning_steps
dtype: string
- name: condition_tier
dtype: string
- name: is_renovated
dtype: bool
- name: is_raw_shell
dtype: bool
- name: year_built_extracted
dtype: int64
- name: logprob_A
dtype: float64
- name: logprob_B
dtype: float64
- name: logprob_C
dtype: float64
- name: logprob_D
dtype: float64
- name: logprob_N
dtype: float64
- name: proc
dtype: int64
splits:
- name: train
num_bytes: 14779851
num_examples: 10152
download_size: 6919567
dataset_size: 14779851
configs:
- config_name: default
data_files:
- split: train
path: data/train-*