metadata
dataset_info:
features:
- name: conv
list:
- name: user
dtype: string
- name: sys
dtype: string
- name: id
dtype: string
- name: do_inference
dtype: bool
- name: inst
dtype: string
- name: key
dtype: int64
- name: prompt
dtype: string
- name: id
dtype: int64
splits:
- name: train
num_bytes: 795330
num_examples: 2000
- name: test
num_bytes: 795330
num_examples: 2000
download_size: 286336
dataset_size: 795330
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/train-*
- config_name: split_prompt
data_files:
- split: test
path: data/train-*