Table Question Answering on ROBUT-WTQ (dev)
61AccuracyOmniTab
Evaluation results
| Method | Links | |
|---|---|---|
| OmniTabEvaluation Protocol=Fine-tuned, Pre-training=Yes2023.06 | 61 | |
| TAPEXEvaluation Protocol=Fine-tuned, Pre-training=Yes, Backbone=BART2023.06 | 57.3 | |
| TableFormerEvaluation Protocol=Fine-tuned, Pre-training=Yes, Backbone=TAPAS-based2023.06 | 51.3 | |
| TAPASEvaluation Protocol=Fine-tuned, Pre-training=Yes, Backbone=BERT2023.06 | 48.3 | |
| GPT-3Evaluation Protocol=Few-shot, Model=text-davinci-003, Samples=2002023.06 | 42.9 |