Table Question Answering on HiTab (dev)
63.51AccuracyTAMA
Evaluation Results
| Method | Links | |
|---|---|---|
| TAMAbackbone=LLaMA 3.1 8B Instruct2025.01 | 63.51 | |
| GPT-42025.01 | 48.4 | |
| GPT-3.52025.01 | 43.62 | |
| LLaMA 3.1 8B Instructstatus=base model2025.01 | 32.83 |
Loading the SOTA2 catalog…