metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
- name: source_file
dtype: string
- name: category
dtype: string
- name: secondary_categories
sequence: string
- name: scenario
dtype: string
- name: language
dtype: string
- name: quality_scores
struct:
- name: completeness
dtype: int64
- name: cot_quality
dtype: int64
- name: embedding_similarity
dtype: int64
- name: factual_correctness
dtype: int64
- name: is_cot
dtype: bool
- name: relevance
dtype: int64
- name: verdict
dtype: string
- name: generator_model
dtype: string
- name: prompt_version
dtype: string
splits:
- name: train
num_bytes: 1919477003
num_examples: 207903
- name: validation
num_bytes: 220844193
num_examples: 23973
download_size: 1503509652
dataset_size: 2140321196
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*