metadata
dataset_info:
features:
- name: model_name
dtype: string
- name: prompt
dtype: string
- name: generated_text
dtype: string
- name: sample_id
dtype: int64
- name: prompt_length
dtype: int64
- name: generated_length
dtype: int64
- name: url
dtype: string
- name: original_instruction
dtype: string
- name: original_input
dtype: string
splits:
- name: train
num_bytes: 15247
num_examples: 5
download_size: 17902
dataset_size: 15247
configs:
- config_name: default
data_files:
- split: train
path: data/train-*