Supervised Fine-Tuning on Marathi Multilingual SFT (test)
34.76Best-k AccuracyTV+KMM
Evaluation results
| Method | Links | |
|---|---|---|
| TV+KMMEvaluation Protocol=SFT, Seeds=52026.05 | 34.76 | |
| One Step+KMMEvaluation Protocol=SFT, Seeds=52026.05 | 34.6 | |
| GradEX-REEvaluation Protocol=SFT, Seeds=52026.05 | 34.53 | |
| One StepEvaluation Protocol=SFT, Seeds=52026.05 | 34.48 | |
| RandomEvaluation Protocol=SFT, Seeds=52026.05 | 34.46 | |
| DataModelEvaluation Protocol=SFT, Seeds=52026.05 | 34.43 | |
| GradEX-FSEvaluation Protocol=SFT, Seeds=52026.05 | 34.41 | |
| TVEvaluation Protocol=SFT, Seeds=52026.05 | 34.37 | |
| DataModel-CSEvaluation Protocol=SFT, Seeds=52026.05 | 33.96 |