GSM8K
Benchmarks
TaskDatasetSOTA resultTrendResults
GSM8K
1,499.6Throughput
18
GSM8K
0.027Brier Score
18
GSM8k
74.04Accuracy
18
GSM8k
39.69Throughput (tokens/s)
18
GSM8K
92.9Score
18
GSM8K
90.9SC Accuracy
18
GSM8K randomly sampled 400 instances (test)
86Accuracy
18
GSM8K 200 examples (test)
81.2Accuracy
18
GSM8K
94.31Accuracy
17
LG-GSM8K
47.33Accuracy (LG-GSM8K)
17
GSM8K
91.5Accuracy
17
GSM8K
14.71AVGLEN
17
GSM8K (test)
5.6Relative Speedup
17
GSM8K (test)
94.62ACC
17
GSM8K
42.38GSM8K Score
17
GSM8K
66.4m1@t1
17
GSM8K (test)
84.5ACC-op
17
GSM8K Aug
83.1Accuracy
16
GSM8k (val)
70.2Accuracy
16
GSM8K
6.02Speedup
16
GSM8K
96.06Accuracy
16
GSM8K trace-level 60/20/20 (test)
88.8Repair Accuracy
16
GSM8K
78Knowledge Recall
16
GSM8K
99.97LLM Trust Score
16
GSM8K original and paraphrased (train)
33.4S_ref
16