Regression on TabArena
1,959Elo RatingTabPFN-3
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| TabPFN-3Evaluation Protocol=Thinking, Train time per 1K [s]=43.00, Predict time per 1K [s]=3.262026.05 | 1,959 | 3.2 | 0.9 | |
| TabPFN-3Evaluation Protocol=Default (D), Train time per 1K [s]=1.69, Predict time per 1K [s]=0.572026.05 | 1,827 | 2.5 | 1.6 | |
| AutoGluon 1.5Evaluation Protocol=extreme, 4h, Train time per 1K [s]=335.03, Predict time per 1K [s]=4.332026.05 | 1,804 | 1.1 | 3.2 | |
| TabPFN-2.6Evaluation Protocol=Default (D), Train time per 1K [s]=8.52, Predict time per 1K [s]=0.702026.05 | 1,776 | 0.8 | 3.3 | |
| RealTabPFN-2.5Evaluation Protocol=Tuned+Ensemble (T+E), Train time per 1K [s]=1709.05, Predict time per 1K [s]=8.122026.05 | 1,774 | 0.5 | 2.6 | |
| TabDPTEvaluation Protocol=Tuned+Ensemble (T+E), Train time per 1K [s]=4786.55, Predict time per 1K [s]=239.542026.05 | 1,748 | 1.5 | 3.5 | |
| TabICLv2Evaluation Protocol=Default (D), Train time per 1K [s]=2.10, Predict time per 1K [s]=0.252026.05 | 1,700 | 1.2 | 3.2 | |
| TabDPTEvaluation Protocol=Tuned (T), Train time per 1K [s]=4786.55, Predict time per 1K [s]=38.502026.05 | 1,696 | 0.1 | 3.9 | |
| RealMLPEvaluation Protocol=Tuned+Ensemble (T+E), Train time per 1K [s]=3995.01, Predict time per 1K [s]=10.052026.05 | 1,677 | 0.2 | 4.3 | |
| RealTabPFN-2.5Evaluation Protocol=Tuned (T), Train time per 1K [s]=1709.05, Predict time per 1K [s]=0.812026.05 | 1,654 | 0.2 | 3.4 | |
| TabDPTEvaluation Protocol=Default (D), Train time per 1K [s]=46.62, Predict time per 1K [s]=39.212026.05 | 1,604 | 0 | 4.9 | |
| RealMLPEvaluation Protocol=Tuned (T), Train time per 1K [s]=3995.01, Predict time per 1K [s]=0.842026.05 | 1,574 | 0 | 5.3 | |
| RealTabPFN-2.5Evaluation Protocol=Default (D), Train time per 1K [s]=7.04, Predict time per 1K [s]=0.512026.05 | 1,558 | 0 | 4.9 | |
| CatBoostEvaluation Protocol=Tuned+Ensemble (T+E), Train time per 1K [s]=3552.96, Predict time per 1K [s]=0.972026.05 | 1,513 | 0 | 7.3 | |
| LightGBMEvaluation Protocol=Tuned+Ensemble (T+E), Train time per 1K [s]=700.15, Predict time per 1K [s]=9.322026.05 | 1,509 | 0 | 7.7 | |
| CatBoostEvaluation Protocol=Tuned (T), Train time per 1K [s]=3552.96, Predict time per 1K [s]=0.102026.05 | 1,489 | 0.1 | 7.4 | |
| TabMEvaluation Protocol=Tuned+Ensemble (T+E), Train time per 1K [s]=4158.29, Predict time per 1K [s]=1.412026.05 | 1,463 | 0 | 6.2 | |
| LightGBMEvaluation Protocol=Tuned (T), Train time per 1K [s]=700.15, Predict time per 1K [s]=0.972026.05 | 1,440 | 0 | 8.3 | |
| XGBoostEvaluation Protocol=Tuned+Ensemble (T+E), Train time per 1K [s]=834.93, Predict time per 1K [s]=2.612026.05 | 1,424 | 0 | 8.2 | |
| XGBoostEvaluation Protocol=Tuned (T), Train time per 1K [s]=834.93, Predict time per 1K [s]=0.392026.05 | 1,403 | 0 | 8.4 | |
| CatBoostEvaluation Protocol=Default (D), Train time per 1K [s]=10.89, Predict time per 1K [s]=0.092026.05 | 1,389 | 0 | 8.9 | |
| TabMEvaluation Protocol=Tuned (T), Train time per 1K [s]=4158.29, Predict time per 1K [s]=0.172026.05 | 1,381 | 0 | 7.1 | |
| TabMEvaluation Protocol=Default (D), Train time per 1K [s]=13.32, Predict time per 1K [s]=0.132026.05 | 1,284 | 0 | 8.8 | |
| RealMLPEvaluation Protocol=Default (D), Train time per 1K [s]=8.90, Predict time per 1K [s]=1.642026.05 | 1,235 | 0 | 9.8 | |
| LightGBMEvaluation Protocol=Default (D), Train time per 1K [s]=2.11, Predict time per 1K [s]=0.272026.05 | 1,210 | 0 | 10.7 | |
| XGBoostEvaluation Protocol=Default (D), Train time per 1K [s]=2.24, Predict time per 1K [s]=0.242026.05 | 1,190 | 0 | 11.3 |