SAVRN
Search Contact SAVRN

Dataset · Text classification

super_glue

by Amanpreet Singh aps/super_glue

SuperGLUE (https://super.gluebenchmark.com/) is a new benchmark styled after GLUE with a new set of more difficult language understanding tasks, improved resources, and a new public leaderboard. An example of 'test' looks as follows.

Rows196,313
Configurations11
Size102.5 MB
Licenseother
AccessPublicly accessible
Monthly Downloads685.9k

Dataset Card

SuperGLUE (https://super.gluebenchmark.com/) is a new benchmark styled after GLUE with a new set of more difficult language understanding tasks, improved resources, and a new public leaderboard. An example of 'test' looks as follows. An example of 'test' looks as follows. An example of 'train' looks as follows. An example of 'train' looks as follows. An example of 'train' looks as follows. The data fields are the same among all splits. - sentence1: a string feature. - sentence2: a string feature. - idx: a int32 feature. - label: a classification label, with possible values including entailment (0), notentailment (1). - premise: a string feature. - hypothesis: a string feature. - idx: a…

Excerpt from the card by Amanpreet Singh, licensed other.

Structure

axb 1,104 rows

SplitRowsSize
test1,104243.8 KB
sentence1stringsentence2stringidxint32labelClassLabel

axg 356 rows

SplitRowsSize
test35653.4 KB
premisestringhypothesisstringidxint32labelClassLabel

boolq 15,942 rows

SplitRowsSize
train9,4276.2 MB
validation3,2702.1 MB
test3,2452.0 MB
questionstringpassagestringidxint32labelClassLabel

cb 556 rows

SplitRowsSize
train25087.1 KB
validation5621.9 KB
test25093.6 KB
premisestringhypothesisstringidxint32labelClassLabel

copa 1,000 rows

SplitRowsSize
train40049.3 KB
validation10012.5 KB
test50060.0 KB
premisestringchoice1stringchoice2stringquestionstringidxint32labelClassLabel

multirc 41,784 rows

SplitRowsSize
train27,24350.9 MB
validation4,8487.8 MB
test9,69315.7 MB
paragraphstringquestionstringanswerstringidxvaluelabelClassLabel

record 120,730 rows

SplitRowsSize
train100,730179.2 MB
validation10,00017.6 MB
test10,00017.2 MB
passagestringquerystringentitiesSequenceentity_spansSequenceanswersSequenceidxvalue

rte 5,767 rows

SplitRowsSize
train2,490857.0 KB
validation27790.8 KB
test3,000724.4 KB
premisestringhypothesisstringidxint32labelClassLabel

wic 7,466 rows

SplitRowsSize
train5,428661.9 KB
validation63882.3 KB
test1,400179.0 KB
wordstringsentence1stringsentence2stringstart1int32start2int32end1int32end2int32idxint32labelClassLabel

wsc 804 rows

SplitRowsSize
train55489.6 KB
validation10421.6 KB
test14631.5 KB
textstringspan1_indexint32span2_indexint32span1_textstringspan2_textstringidxint32labelClassLabel

wsc.fixed 804 rows

SplitRowsSize
train55489.6 KB
validation10421.6 KB
test14631.5 KB
textstringspan1_indexint32span2_indexint32span1_textstringspan2_textstringidxint32labelClassLabel

Details

Repository
aps/super_glue
Publisher
Amanpreet Singh
Task category
Text classification
Tags
superglue, NLU, natural language understanding
Size category
10K<n<100K
Languages
en
Revision
3de24cf8022e94f4ee4b9d55a6f539891524d646
Last updated
2025-05-16

SAVRN's Notes on super_glue

SuperGLUE carries 196,313 rows across 11 configs in 102.5 MB, license listed as other, 685.9k monthly downloads, under the super_glue name with NLU tags. BoolQ is the biggest config at 15,942 rows; the diagnostic AX-b and AX-g sets are 1,104 and 356 rows. Our page shows two inbound models, both DeBERTa long-NLI variants: this is benchmark territory, not training territory.

The publisher line names an individual mirror, Amanpreet Singh's aps/super_glue, rather than the benchmark consortium, so pin the revision, last updated 2025-05-16, if you need reproducibility. Same play as GLUE: train on the larger configs' train splits, never on the eval sets, and read the per-task terms since other bundles several licenses. Expect modern base models to post near-ceiling scores.

SAVRN Research

Files

31 files, 102.5 MB in total.

Data29 files · 102.5 MB
Documentation1 file · 19.7 KB
Repository1 file · 1.2 KB
Every file
FileTypeSizeSHA-256
axb/test-00000-of-00001.parquetData80.9 KB655cadf2b23e
axg/test-00000-of-00001.parquetData14.3 KB4ed3eddc7286
boolq/test-00000-of-00001.parquetData1.3 MB95a74cd7c8fd
boolq/train-00000-of-00001.parquetData3.8 MB41e2ce07ac0d
boolq/validation-00000-of-00001.parquetData1.3 MBf46438b8cce0
cb/test-00000-of-00001.parquetData63.5 KBe9b1164fed6c
cb/train-00000-of-00001.parquetData58.0 KBa338d1276fd0
cb/validation-00000-of-00001.parquetData18.0 KB36cf71dba8a3
copa/test-00000-of-00001.parquetData40.2 KB7d7d6e12adc0
copa/train-00000-of-00001.parquetData33.9 KB8e243c3a2428
copa/validation-00000-of-00001.parquetData12.0 KB247a7f586c5c
multirc/test-00000-of-00001.parquetData580.9 KB465107b93b83
multirc/train-00000-of-00001.parquetData1.7 MBced6d731a335
multirc/validation-00000-of-00001.parquetData306.4 KB5eb64893afc5
record/test-00000-of-00001.parquetData8.0 MB6a91cd50dbdf
record/train-00000-of-00001.parquetData75.1 MB20ae20deb498
record/validation-00000-of-00001.parquetData8.1 MB3662aa327fec
rte/test-00000-of-00001.parquetData622.0 KB5a4c57d7b81e
rte/train-00000-of-00001.parquetData586.2 KB16c83717da2e
rte/validation-00000-of-00001.parquetData69.8 KBb3740aac8052
wic/test-00000-of-00001.parquetData126.8 KBe77d12183550
wic/train-00000-of-00001.parquetData410.3 KBd6d2231730f5
wic/validation-00000-of-00001.parquetData60.1 KB22cc5df62bf7
wsc.fixed/test-00000-of-00001.parquetData12.2 KB5308e3ca6543
wsc.fixed/train-00000-of-00001.parquetData27.3 KB13086042694c
wsc.fixed/validation-00000-of-00001.parquetData10.2 KBaa64cc8f27c7
wsc/test-00000-of-00001.parquetData12.2 KB4a39c4f576a4
wsc/train-00000-of-00001.parquetData27.2 KBa1fb4be43c74
wsc/validation-00000-of-00001.parquetData10.2 KB1459bd736725
README.mdDocumentation19.7 KB—
.gitattributesRepository1.2 KB—

License and Download

License
other
Access
No access gate
Download from Amanpreet Singh

Released by Amanpreet Singh through its official repository on Hugging Face.

Models Trained on This Dataset