Skip to content

Dataset · mteb

STS12 semantic textual similarity benchmark

mteb/sts12-sts

Classic dataset for evaluating text embedding quality from the MTEB benchmark.

What's inside

Includes between 1 and 10 thousand English examples from the SemEval-2012 Task 6 challenge, integrated into the popular MTEB benchmark. The data covers encyclopedic articles, news, and reference literature, and is designed for evaluating semantic sentence similarity.

What you can do with it

  • Testing and comparing text embedding models within the MTEB framework
  • Calibrating semantic search and document ranking systems
Tasks
Sentence similarity
Languages
English
Size
1K–10K rows
Downloads this month
48.4K
Likes
15

How to load it

from datasets import load_dataset

ds = load_dataset("mteb/sts12-sts")