Split and Rephrase on BiSECT (test)
76.32BESCpythia-6.9B-BI
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| pythia-6.9B-BIBackbone=Pythia-6.9B, Training samples=30k, Fine-tuned=true2023.12 | 76.32 | 42.4 | 45.06 | 16.81 | 65.93 | 52.45 | 95.17 | — | — | — | — | |
| pythia-6.9B-BIBackbone=Pythia-6.9B, Training samples=6K, Fine-tuned=true2023.12 | 76.12 | 41.3 | 38.93 | 12.92 | 63.59 | 40.27 | 81.55 | — | — | — | — | |
| alpaca-7BBackbone=Alpaca-7B, Mode=few-shot, Few-shot Context=BiSECT2023.12 | 73.17 | 36.8 | 35.82 | 8.57 | 60.46 | 38.43 | 65.86 | — | — | — | — | |
| BiSECT seq2seqModel=BiSECT seq2seq2023.12 | 72.51 | 36.8 | 33.21 | 7.34 | 59.91 | 32.4 | 99.48 | — | — | — | — | |
| GPT-4 TurboModel=GPT-4 Turbo2023.12 | 72.04 | 34.7 | 37.18 | 9.65 | 60.33 | 41.55 | 100 | — | — | — | — | |
| alpaca-7BBackbone=Alpaca-7B, Mode=zero-shot2023.12 | 69.32 | 36.2 | 29.82 | 5.08 | 60.39 | 23.99 | 88.62 | — | — | — | — | |
| Pythia-6.9B-DEAlternative Method (S)=BiSECT2023.12 | — | — | — | — | — | — | — | 43.09 | 38.45 | 18.45 | 0.49 |