List your company

Math Problem Solving on MWPBENCH (out-of-domain)

24.8College Math AccWizardMath-Mistral-RL
0.2566.6281319.372Aug 18, 2023

Evaluation results

MethodLinks
WizardMath-Mistral-RLBackbone=Mistral 7B, Training Stage=RL2023.0824.844.871.252.637.224.564.745.7
GPT-4Model Category=Proprietary models2023.0824.451.876.561.535.428.268.649.5
WizardMath-Mistral-SFTBackbone=Mistral 7B, Training Stage=SFT2023.0824.342.766.649.735.222.763.143.5
WizardMath-RLBackbone=LLaMA-2 13B, Training Stage=RL2023.0822.943.370.350.833.125.764.744.4
WizardMath-SFTBackbone=LLaMA-2 13B, Training Stage=SFT2023.0822.242.565.947.631.623.559.741.9
MathScale MistralBackbone=Mistral 7B2023.0821.839.964.44621.414.357.837.9
GPT-3.5-TurboModel Category=Proprietary models2023.0821.642.962.54423.215.355.837.9
WizardMath-RLBackbone=LLaMA-2 7B, Training Stage=RL2023.0821.240.267.346.128.918.762.740.7
WizardMath-SFTBackbone=LLaMA-2 7B, Training Stage=SFT2023.0821.138.562.443.826.317.758.338.3
MathScale 7BBackbone=LLaMA-2 7B2023.0820.935.25941.819.612.657.835.3
MathScale 13BBackbone=LLaMA-2 13B2023.0820.438.161.143.72012.355.835.9
DART-Math-Prop2DiffBackbone=Mistral 7B2023.0819.937.462.244.927.218.162.738.9
DART-Math-UniformBackbone=Mistral 7B2023.0819.434.861.644.82716.159.837.6
MetaMath MistralBackbone=Mistral 7B2023.0815.731.455.138.115.310.150.930.9
Xwin-Math-V1.1Backbone=LLaMA-2 7B2023.0814.929.759.640.815.98.45131.5
DART-Math-UniformBackbone=LLaMA-2 7B2023.081227.347.932.914.811.145.127.3
DART-Math-Prop2DiffBackbone=LLaMA-2 7B2023.0811.927.749.934.312.810.647.127.8
MetaMathBackbone=LLaMA-2 13B2023.0810.125.448.631.69.65.638.224.2
MetaMathBackbone=LLaMA-2 7B2023.089.422.54429.95.95.136.221.9
GAIR-AbelBackbone=LLaMA-2 13B2023.087.921.142.227.874.930.320.2
Mistral 7BBackbone=Mistral 7B2023.087.517.918.515.56.25.922.513.4
GAIR-AbelBackbone=LLaMA-2 7B2023.086.618.335.424.54.34.423.516.7
MAmmoTH-CoTBackbone=LLaMA-2 13B2023.086.517.339.528.15.94.920.517.5
MAmmoTH-CoTBackbone=LLaMA-2 7B2023.086.213.334.621.43.92.719.614.5
LLaMA-2 7BBackbone=LLaMA-2 7B2023.082.37.66.87.32.12.92.94.6
LLaMA-2 13BBackbone=LLaMA-2 13B2023.081.26.39.57.90.70.46.84.7