List your company

Multilingual Mathematical Reasoning on MGSM8KInstruct (test)

47.6Accuracy (En)LM Base Model
16.81624.80832.840.792Feb 24, 2025

Evaluation results

MethodLinks
LM Base ModelLM Component=Included, Math Component=Excluded, Merging Method=None2025.0247.61634.822.434.826.434.823.234.425.229.9
Math Base ModelLM Component=Excluded, Math Component=Included, Merging Method=None2025.0243.22.813.212.821.230.820.41018.412.818.5
LED-MergingLM Component=Included, Math Component=Included2025.0236.87.627.224.820.422.418.418.821.216.421.3
Ties-MergingLM Component=Included, Math Component=Included2025.0229.69.223.615.621.619.618.417.623.213.219.2
Task ArithmeticLM Component=Included, Math Component=Included2025.02281017.212.817.213.213.214.4181415.8
BreadcrumbsLM Component=Included, Math Component=Included2025.0222.87.22.813.613.213.612.89.615.61212.3
Model StockLM Component=Included, Math Component=Included2025.02188.417.61012.411.613.218.418.41213.9