List your company
ResearchFormerly WizWand

Mathematical Reasoning on GSM8K (Spd, tau)

7.46SpdGOOSE
0.74162.48584.235.9742Apr 2, 2026
Updated 7mo ago

Evaluation results

MethodLinks
GOOSEModel=Llama-3-8B, GPU=NVIDIA A402026.047.469.25
PLDModel=Llama-3-8B, GPU=NVIDIA A402026.044.044.18
TRModel=Llama-3-8B, GPU=NVIDIA A402026.043.684.52
GOOSEModel=Qwen2-8B, GPU=NVIDIA A402026.043.233.87
EAGLE-2Model=Vicuna-13B, GPU=NVIDIA A402026.043.184.98
EAGLE-2Model=Llama-3-8B, GPU=NVIDIA A402026.042.755.52
EAGLE-2Model=Vicuna-7B, GPU=NVIDIA A402026.042.734.51
GOOSEModel=Vicuna-7B, GPU=NVIDIA A402026.042.673.23
GOOSEModel=Vicuna-13B, GPU=NVIDIA A402026.042.663.14
TRModel=Qwen2-8B, GPU=NVIDIA A402026.042.63.03
LookaheadModel=Llama-3-8B, GPU=NVIDIA A402026.042.592.8
TRModel=Vicuna-13B, GPU=NVIDIA A402026.042.563.02
GOOSEModel=Vicuna-33B, GPU=NVIDIA A1002026.042.533.1
TRModel=Vicuna-7B, GPU=NVIDIA A402026.042.462.96
TRModel=Vicuna-33B, GPU=NVIDIA A1002026.042.142.63
LookaheadModel=Qwen2-8B, GPU=NVIDIA A402026.041.931.88
PLDModel=Qwen2-8B, GPU=NVIDIA A402026.041.651.71
LookaheadModel=Vicuna-7B, GPU=NVIDIA A402026.041.631.67
LookaheadModel=Vicuna-13B, GPU=NVIDIA A402026.041.581.62
LookaheadModel=Vicuna-33B, GPU=NVIDIA A1002026.041.491.55
PLDModel=Vicuna-7B, GPU=NVIDIA A402026.041.431.46
PLDModel=Vicuna-13B, GPU=NVIDIA A402026.041.331.36
PLDModel=Vicuna-33B, GPU=NVIDIA A1002026.041.291.33
RESTModel=Qwen2-8B, GPU=NVIDIA A402026.041.211.16
RESTModel=Vicuna-13B, GPU=NVIDIA A402026.041.181.21
RESTModel=Vicuna-7B, GPU=NVIDIA A402026.041.171.19
RESTModel=Vicuna-33B, GPU=NVIDIA A1002026.041.141.18
RESTModel=Llama-3-8B, GPU=NVIDIA A402026.041.061.07
ARModel=Vicuna-7B, GPU=NVIDIA A402026.0411
ARModel=Llama-3-8B, GPU=NVIDIA A402026.0411
ARModel=Qwen2-8B, GPU=NVIDIA A402026.0411
ARModel=Vicuna-13B, GPU=NVIDIA A402026.0411
ARModel=Vicuna-33B, GPU=NVIDIA A1002026.0411