Datasets:
metadata
language:
- en
task_categories:
- text-classification
dataset_info:
- config_name: sst5
features:
- name: id
dtype: int64
- name: phrase
dtype: string
- name: sentiment
dtype: float32
- name: label
dtype:
class_label:
names:
'0': very negative
'1': negative
'2': neutral
'3': positive
'4': very positive
- config_name: sst2
features:
- name: id
dtype: int64
- name: phrase
dtype: string
- name: sentiment
dtype: float32
- name: label
dtype:
class_label:
names:
'0': negative
'1': positive
configs:
- config_name: sst5
data_files:
- split: complete
path: sst5.csv
default: true
- config_name: sst2
data_files:
- split: complete
path: sst2.csv