LetTokenBurnQuora Question Pairs (QQP) dataset consists of over 400,000 question pairs, and each question pair is annotated with a binary value indicating whether the two questions are paraphrase of each other.
$ npm i -g lettokenburn-cli$ ltb bench info quora-question-pairs
External harness. No Harbor adapter has been generated for this benchmark yet. `ltb bench info` prints the known links.
Runs happen on your machine with your own subscription; only the score summary is pushed. CLI docs
| Harness | Source | Confidence | Link | |||
|---|---|---|---|---|---|---|
| T5-11B | 90.4 | N/A | Independent | medium | Oct 23, 2019 | arxiv.org |
| ERNIE 2.0 Large | 90.1 | N/A | Independent | medium | Jul 29, 2019 | arxiv.org |
| ERNIE 2.0 Base | 89.8 | N/A | Independent | medium | Jul 29, 2019 | arxiv.org |
| ERNIE | 71.2 | N/A | Independent | medium | May 17, 2019 | arxiv.org |
Every number links to where it was published. Low-confidence rows are muted and need per-cell verification.
Quora Question Pairs (QQP) dataset consists of over 400,000 question pairs, and each question pair is annotated with a binary value indicating whether the two questions are paraphrase of each other.
Source: [Bilateral Multi-Perspective Matching for Natural Language Sentences](/paper/bilateral-multi-perspective-matching-for)