Table 1 Summary of dataset size compared to KLUE and GLUE tasks.

From: Multi-task learning approach for utilizing temporal relations in natural language understanding tasks

Language

Task

# Train

# Valid

# Test

Korean

STS

9335

519

2333

NLI

19,999

3000

4999

NER

16,807

5000

4201

TLINK-C

1804

450

250

English

STS

5749

1500

1379

NLI

549,367

9842

9824

NER

14,041

3250

3453

TLINK-C

14,056

4016

2009

Back to article page