Language-to-Code Generation on WikiTQ official (test)
74.6Execution AccuracyOracle
Evaluation results
| Method | Links | |
|---|---|---|
| Oraclegenerator_model=code-davinci-0022023.02 | 74.6 | |
| LEVERgenerator_model=code-davinci-0022023.02 | 62.9 | |
| OmniTabfinetuning=true2023.02 | 62.8 | |
| Codex Binderfinetuning=false2023.02 | 61.9 | |
| TaCubefinetuning=true2023.02 | 59.6 | |
| TaPEX*finetuning=true2023.02 | 57.5 | |
| Codex SQLfinetuning=false2023.02 | 55.1 | |
| EP + Votinggenerator_model=code-davinci-0022023.02 | 53.6 | |
| EP + MLgenerator_model=code-davinci-0022023.02 | 52.5 | |
| Greedygenerator_model=code-davinci-0022023.02 | 50.9 | |
| MLgenerator_model=code-davinci-0022023.02 | 50.9 | |
| Codex QAfinetuning=false2023.02 | 47.6 |