metadata
dataset_info:
features:
- name: instruction
dtype: string
- name: Yurts_RAG_context
dtype: string
- name: num_contexts
dtype: int64
- name: max_context_size
dtype: int64
- name: context4instruction
dtype: string
- name: pdf_path
dtype: string
- name: num_tokens
dtype: int64
- name: Llama-3 response
dtype: string
- name: GPT-4o response
dtype: string
- name: Claude response
dtype: 'null'
- name: likert_score_with_real_instruction
dtype: int64
splits:
- name: train
num_bytes: 660248495
num_examples: 2065
download_size: 106973303
dataset_size: 660248495
configs:
- config_name: default
data_files:
- split: train
path: data/train-*