Question 1

What are Paraphrasing Identification datasets used for?

Accepted Answer

Paraphrasing Identification datasets are collections of labelled or raw data used to train, fine-tune, and evaluate models on the paraphrasing identification task. This page lists 4 such datasets, each linking to its source and paper.

Question 2

Which Paraphrasing Identification dataset is best for benchmarking?

Accepted Answer

1 of these Paraphrasing Identification datasets are tracked as benchmarks, including Microsoft Research Paraphrase Corpus (MRPC). See the Benchmarks section for model leaderboards.

Question 3

How many Paraphrasing Identification datasets are there?

Accepted Answer

We catalog 4 Paraphrasing Identification datasets in one searchable directory.

Paraphrasing Identification Datasets

What languages do paraphrasing identification datasets cover?

Explore other dataset tasks

Frequently asked questions