metadata
dataset_info:
features:
- name: input
dtype: string
- name: output
dtype: int64
- name: instruction
dtype: string
- name: sentiment_name
dtype: string
splits:
- name: train
num_bytes: 18115769
num_examples: 78045
- name: validation
num_bytes: 4308552
num_examples: 18743
download_size: 7821945
dataset_size: 22424321
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*