GSM8K
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
gsm8k 1.0 (test)
96.19Overall Accuracy
6
GSM8K
33.07Selection Accuracy
6
GSM8K (test)
56.3Accuracy
6
GSM8K (test)
0.749Accuracy
6
GSM8K
0.683Accuracy
6
GSM8K (test)
86Accuracy
6
GSM8K (test)
85.81Accuracy
6
GSM8K
96.4GSM8K Pass@1 Accuracy
6
GSM8K
79.68Accuracy
5
GSM8K @1 (test)
76.42Accuracy
5
GSM8K (val)
112Throughput (tok/s)
5
GSM8K
93.21Accuracy (Source Lib.)
5
GSM8K
79.38Accuracy (GSM8K)
5
GSM8K
1,163Latency (NoKV) (ms)
5
GSM8K
71.9T Score
5
GSM8K
99Threshold Accuracy
5
GSM8K paired (evaluation)
0.926AUROC
5
GSM8K
53.55TPS
5
GSM8K
76.96Accuracy
5
GSM8k
86.7Accuracy
5
GSM8k (test)
91.3Accuracy
5
GSM8K
81.97Accuracy (GSM8K)
5
GSM8K
7.98Average Acceptance Length (τ)
5
GSM8K
73.8Accuracy
5
GSM8K
100Accuracy
5