NL-to-Scala on Combined NL-to-Scala Benchmark (WikiSQL, WikiTQ, Text2Analysis, Spider, BIRD)
49.1Quality ScoreTable-Specialist
Evaluation results
| Method | Links | |
|---|---|---|
| Table-SpecialistFine-tuning Strategy=Specialist, Base Model=GPT-3.52024.10 | 49.1 | |
| TableGPTFine-tuning Strategy=Generalist, Base Model=GPT-3.52024.10 | 30.4 | |
| TableLlamaFine-tuning Strategy=Generalist, Base Model=GPT-3.52024.10 | 27.9 |