Counterfactual Story Generation on Counterfactual Story (test)
87.18BLEU-4ClarET
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ClarET2022.03 | 87.18 | 80.74 | 81.48 | |
| ClarETSize=400M, Mode=Fine-tuned2022.03 | 87.18 | 80.74 | 81.48 | |
| BART-largemodel_size=large2022.03 | 82.91 | 76.44 | 79.5 | |
| BARTSize=400M, Mode=Fine-tuned2022.03 | 82.91 | 76.44 | 79.5 | |
| GPT2-M + Rec + CF2022.03 | 75.92 | 70.93 | 62.49 | |
| GLM-largemodel_size=large2022.03 | 75.81 | 70.03 | 68.23 | |
| GLMSize=335M, Mode=Fine-tuned2022.03 | 75.81 | 70.03 | 68.23 | |
| GPT2-M + Sup2022.03 | 75.71 | 72.72 | 62.39 | |
| GPT2-MSize=335M, Mode=Fine-tuned2022.03 | 75.71 | 72.72 | 62.39 | |
| GPT2-S2022.03 | 69.27 | 65.72 | 60.53 | |
| GPT2-SSize=124M, Mode=Fine-tuned2022.03 | 69.27 | 65.72 | 60.53 | |
| Human2022.03 | 64.93 | 67.64 | 61.87 | |
| ClarETZero-shot=true2022.03 | 23.75 | 43.03 | 63.93 | |
| DELOREANZero-shot=true2022.03 | 21.35 | 40.73 | 63.36 | |
| BART-largeZero-shot=true2022.03 | 7.08 | 30.6 | 61.58 | |
| Zero-Shot-RankedZero-shot=true2022.03 | 2.26 | 25.81 | 60.07 | |
| GPT2-MZero-shot=true2022.03 | 1.51 | 19.41 | 60.17 | |
| GPT2-SZero-shot=true2022.03 | 1.28 | 20.27 | 59.62 | |
| GPTZero-shot=true2022.03 | 1.25 | 18.26 | 59.5 |