PierreLepagnol/WRENCH

---
task_categories:
- text-classification
- token-classification
size_categories:
- 10K<n<100K
dataset_info:
- config_name: yelp
features:
- name: text
dtype: string
- name: label
dtype: int8
- name: weak_labels
sequence: int8
splits:
- name: train
num_bytes: 22618599
num_examples: 30400
- name: validation
num_bytes: 2824249
num_examples: 3800
- name: test
num_bytes: 2709033
num_examples: 3800
download_size: 37356054
dataset_size: 28151881
- config_name: imdb
features:
- name: text
dtype: string
- name: label
dtype: int8
- name: weak_labels
sequence: int8
splits:
- name: train
num_bytes: 25515241
num_examples: 20000
- name: validation
num_bytes: 3269130
num_examples: 2500
- name: test
num_bytes: 3151954
num_examples: 2500
download_size: 33910706
dataset_size: 31936325
- config_name: agnews
features:
- name: text
dtype: string
- name: label
dtype: int8
- name: weak_labels
sequence: int8
splits:
- name: train
num_bytes: 20357121
num_examples: 96000
- name: validation
num_bytes: 2487983
num_examples: 12000
- name: test
num_bytes: 2521518
num_examples: 12000
download_size: 39149014
dataset_size: 25366622
- config_name: cdr
features:
- name: text
dtype: string
- name: label
dtype: int8
- name: entity1
dtype: string
- name: entity2
dtype: string
- name: weak_labels
sequence: int8
splits:
- name: train
num_bytes: 2318545
num_examples: 8430
- name: validation
num_bytes: 246252
num_examples: 920
- name: test
num_bytes: 1229627
num_examples: 4673
download_size: 11036213
dataset_size: 3794424
- config_name: chemprot
features:
- name: text
dtype: string
- name: label
dtype: int8
- name: entity1
dtype: string
- name: entity2
dtype: string
- name: weak_labels
sequence: int8
splits:
- name: train
num_bytes: 3474637
num_examples: 12861
- name: validation
num_bytes: 435850
num_examples: 1607
- name: test
num_bytes: 434031
num_examples: 1607
download_size: 15743249
dataset_size: 4344518
- config_name: semeval
features:
- name: text
dtype: string
- name: label
dtype: int8
- name: entity1
dtype: string
- name: entity2
dtype: string
- name: weak_labels
sequence: int8
splits:
- name: train
num_bytes: 532785
num_examples: 1749
- name: validation
num_bytes: 54373
num_examples: 178
- name: test
num_bytes: 184826
num_examples: 600
download_size: 2295058
dataset_size: 771984
- config_name: sms
features:
- name: text
dtype: string
- name: label
dtype: int8
- name: weak_labels
sequence: int8
splits:
- name: train
num_bytes: 741520
num_examples: 4571
- name: validation
num_bytes: 81747
num_examples: 500
- name: test
num_bytes: 80152
num_examples: 500
download_size: 6715435
dataset_size: 903419
- config_name: spouse
features:
- name: text
dtype: string
- name: label
dtype: int8
- name: entity1
dtype: string
- name: entity2
dtype: string
- name: weak_labels
sequence: int8
splits:
- name: train
num_bytes: 7550613
num_examples: 22254
- name: validation
num_bytes: 952523
num_examples: 2811
- name: test
num_bytes: 876804
num_examples: 2701
download_size: 22017644
dataset_size: 9379940
- config_name: trec
features:
- name: text
dtype: string
- name: label
dtype: int8
- name: weak_labels
sequence: int8
splits:
- name: train
num_bytes: 610244
num_examples: 4965
- name: validation
num_bytes: 61048
num_examples: 500
- name: test
num_bytes: 56479
num_examples: 500
download_size: 2277067
dataset_size: 727771
- config_name: youtube
features:
- name: text
dtype: string
- name: label
dtype: int8
- name: weak_labels
sequence: int8
splits:
- name: train
num_bytes: 180736
num_examples: 1586
- name: validation
num_bytes: 14659
num_examples: 120
- name: test
num_bytes: 33347
num_examples: 250
download_size: 759494
dataset_size: 228742
---

原始信息汇总

数据集概述

数据集配置及特征

配置名称	特征名称	数据类型
yelp	text	string
yelp	label	int8
yelp	weak_labels	sequence: int8
imdb	text	string
imdb	label	int8
imdb	weak_labels	sequence: int8
agnews	text	string
agnews	label	int8
agnews	weak_labels	sequence: int8
cdr	text	string
cdr	label	int8
cdr	entity1	string
cdr	entity2	string
cdr	weak_labels	sequence: int8
chemprot	text	string
chemprot	label	int8
chemprot	entity1	string
chemprot	entity2	string
chemprot	weak_labels	sequence: int8
semeval	text	string
semeval	label	int8
semeval	entity1	string
semeval	entity2	string
semeval	weak_labels	sequence: int8
sms	text	string
sms	label	int8
sms	weak_labels	sequence: int8
spouse	text	string
spouse	label	int8
spouse	entity1	string
spouse	entity2	string
spouse	weak_labels	sequence: int8
trec	text	string
trec	label	int8
trec	weak_labels	sequence: int8
youtube	text	string
youtube	label	int8
youtube	weak_labels	sequence: int8

数据集大小及分割

配置名称	分割类型	示例数量	字节数
yelp	train	30400	22618599
yelp	validation	3800	2824249
yelp	test	3800	2709033
imdb	train	20000	25515241
imdb	validation	2500	3269130
imdb	test	2500	3151954
agnews	train	96000	20357121
agnews	validation	12000	2487983
agnews	test	12000	2521518
cdr	train	8430	2318545
cdr	validation	920	246252
cdr	test	4673	1229627
chemprot	train	12861	3474637
chemprot	validation	1607	435850
chemprot	test	1607	434031
semeval	train	1749	532785
semeval	validation	178	54373
semeval	test	600	184826
sms	train	4571	741520
sms	validation	500	81747
sms	test	500	80152
spouse	train	22254	7550613
spouse	validation	2811	952523
spouse	test	2701	876804
trec	train	4965	610244
trec	validation	500	61048
trec	test	500	56479
youtube	train	1586	180736
youtube	validation	120	14659
youtube	test	250	33347

下载与数据集大小

配置名称	下载大小	数据集大小
yelp	37356054	28151881
imdb	33910706	31936325
agnews	39149014	25366622
cdr	11036213	3794424
chemprot	15743249	4344518
semeval	2295058	771984
sms	6715435	903419
spouse	22017644	9379940
trec	2277067	727771
youtube	759494	228742