List your company

Accuracy (%) on MMLU

52.15AccuracySFT
28.760434.832740.90546.9773Jul 20, 2025

Evaluation results

MethodLinks
SFTModel=Qwen3-8B-Base2025.0752.15
SFT+GRPOModel=Qwen3-8B-Base2025.0750.85
SFT+PenaltyModel=Qwen3-8B-Base2025.0750.74
SASFTModel=Qwen3-8B-Base2025.0750.09
SFT+PenaltyModel=Gemma-2-9B2025.0746.39
SASFTModel=Gemma-2-9B2025.0745.91
SFTModel=Gemma-2-9B2025.0744.31
SFT+GRPOModel=Gemma-2-9B2025.0744.21
SASFTModel=Qwen3-1.7B-Base2025.0738.38
SFT+GRPOModel=Qwen3-1.7B-Base2025.0737.8
SFT+PenaltyModel=Qwen3-1.7B-Base2025.0737.78
SFTModel=Qwen3-1.7B-Base2025.0737.47
SASFTModel=Llama-3.1-8B2025.0733.12
SFT+PenaltyModel=Gemma-2-2B2025.0730.81
SASFTModel=Gemma-2-2B2025.0730.24
SFTModel=Llama-3.1-8B2025.0729.99
SFTModel=Gemma-2-2B2025.0729.88
SFT+PenaltyModel=Llama-3.1-8B2025.0729.7
SFT+GRPOModel=Llama-3.1-8B2025.0729.67
SFT+GRPOModel=Gemma-2-2B2025.0729.66