Table Question Answering on ROBUT-WTQ Column Adding (test)
81.4Accuracy (Pre-Perturbation)TAPEX
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| TAPEXEvaluation Protocol=Fine-tuned2023.06 | 81.4 | 64.9 | 71 | |
| OmniTabEvaluation Protocol=Fine-tuned2023.06 | 79.7 | 66.7 | 71.3 | |
| GPT-3Evaluation Protocol=Few-shot, Sample Size=2002023.06 | 78.3 | 70.5 | 81 | |
| TAPASEvaluation Protocol=Fine-tuned2023.06 | 77.8 | 69.6 | 81.6 | |
| TableFormerEvaluation Protocol=Fine-tuned2023.06 | 75.4 | 67.3 | 80.1 | |
| OmniTabEvaluation Protocol=Fine-tuned2023.06 | 61.6 | 57.2 | 84.8 | |
| TAPEXEvaluation Protocol=Fine-tuned2023.06 | 57.4 | 50.4 | 80.1 | |
| TableFormerEvaluation Protocol=Fine-tuned2023.06 | 51.9 | 48.7 | 83.5 | |
| TAPASEvaluation Protocol=Fine-tuned2023.06 | 48.9 | 47.1 | 89.3 | |
| GPT-3Evaluation Protocol=Few-shot2023.06 | 41.3 | 36.8 | 85.6 |