Table Question Answering on ROBUT-WTQ Sentence-Level Paraphrase (test)
81.6Accuracy (Pre-perturbation)TAPEX
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| TAPEXEvaluation Protocol=Fine-tuned2023.06 | 81.6 | 68.7 | 80.8 | |
| OmniTabEvaluation Protocol=Fine-tuned2023.06 | 80.6 | 70.1 | 81.3 | |
| TAPASEvaluation Protocol=Fine-tuned2023.06 | 79.3 | 66.7 | 78.6 | |
| TableFormerEvaluation Protocol=Fine-tuned2023.06 | 76.8 | 64.5 | 79.1 | |
| GPT-3Evaluation Protocol=Few-shot, Sample Size=2002023.06 | 75 | 72.6 | 95.1 | |
| OmniTabEvaluation Protocol=Fine-tuned2023.06 | 58.2 | 55.4 | 87 | |
| TAPEXEvaluation Protocol=Fine-tuned2023.06 | 54.8 | 49.5 | 84 | |
| TableFormerEvaluation Protocol=Fine-tuned2023.06 | 49.6 | 44 | 77.1 | |
| TAPASEvaluation Protocol=Fine-tuned2023.06 | 45.6 | 41.1 | 80.8 | |
| GPT-3Evaluation Protocol=Few-shot2023.06 | 41 | 40.5 | 94.2 |