Semantic Textual Similarity On Senteval
評価指標
MRPC
SICK-E
SICK-R
STS
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
モデル名 | MRPC | SICK-E | SICK-R | STS | Paper Title | Repository |
|---|---|---|---|---|---|---|
| Snorkel MeTaL(ensemble) | 91.5/88.5 | - | - | 90.1/89.7* | Training Complex Models with Multi-Task Weak Supervision | |
| XLNet-Large | 93.0/90.7 | - | - | 91.6/91.1* | XLNet: Generalized Autoregressive Pretraining for Language Understanding | |
| GenSen | 78.6/84.4 | 87.8 | 0.888 | 78.9/78.6 | Learning General Purpose Distributed Sentence Representations via Large Scale Multi-task Learning | |
| TF-KLD | 80.4/85.9 | - | - | - | - | - |
| MT-DNN-ensemble | 92.7/90.3 | - | - | 91.1/90.7* | Improving Multi-Task Deep Neural Networks via Knowledge Distillation for Natural Language Understanding | |
| InferSent | 76.2/83.1 | 86.3 | 0.884 | 75.8/75.5 | Supervised Learning of Universal Sentence Representations from Natural Language Inference Data |
0 of 6 row(s) selected.