lm1-misc-pile
/
1b12b12b
/1b112b12bpile
/evaluation
/lm1-1b1-12b-results_lm-eval_global_step23189_2023-01-24-13-53-32_2shots.json
{ | |
"results": { | |
"anli_r1": { | |
"acc": 0.317, | |
"acc_stderr": 0.014721675438880229 | |
}, | |
"anli_r2": { | |
"acc": 0.334, | |
"acc_stderr": 0.01492201952373296 | |
}, | |
"anli_r3": { | |
"acc": 0.3325, | |
"acc_stderr": 0.013605417345710526 | |
}, | |
"cb": { | |
"acc": 0.35714285714285715, | |
"acc_stderr": 0.0646095738380922, | |
"f1": 0.2563323201621074 | |
}, | |
"copa": { | |
"acc": 0.65, | |
"acc_stderr": 0.0479372485441102 | |
}, | |
"hellaswag": { | |
"acc": 0.29685321649073887, | |
"acc_stderr": 0.004559375835805974, | |
"acc_norm": 0.32593108942441745, | |
"acc_norm_stderr": 0.0046776374633913965 | |
}, | |
"rte": { | |
"acc": 0.5234657039711191, | |
"acc_stderr": 0.03006330041190266 | |
}, | |
"winogrande": { | |
"acc": 0.5430149960536701, | |
"acc_stderr": 0.01400038676159829 | |
}, | |
"storycloze_2016": { | |
"acc": 0.584179583110636, | |
"acc_stderr": 0.01139738683314014 | |
}, | |
"boolq": { | |
"acc": 0.4792048929663609, | |
"acc_stderr": 0.008737488341370727 | |
}, | |
"arc_easy": { | |
"acc": 0.47095959595959597, | |
"acc_stderr": 0.01024246382639562, | |
"acc_norm": 0.4494949494949495, | |
"acc_norm_stderr": 0.010207308833916035 | |
}, | |
"arc_challenge": { | |
"acc": 0.20477815699658702, | |
"acc_stderr": 0.011792544338513398, | |
"acc_norm": 0.2431740614334471, | |
"acc_norm_stderr": 0.012536554144587089 | |
}, | |
"sciq": { | |
"acc": 0.841, | |
"acc_stderr": 0.011569479368271289, | |
"acc_norm": 0.819, | |
"acc_norm_stderr": 0.01218143617917791 | |
}, | |
"piqa": { | |
"acc": 0.6392818280739935, | |
"acc_stderr": 0.011204064809088518, | |
"acc_norm": 0.6360174102285092, | |
"acc_norm_stderr": 0.011225875703487171 | |
} | |
}, | |
"versions": { | |
"anli_r1": 0, | |
"anli_r2": 0, | |
"anli_r3": 0, | |
"cb": 1, | |
"copa": 0, | |
"hellaswag": 0, | |
"rte": 0, | |
"winogrande": 0, | |
"storycloze_2016": 0, | |
"boolq": 1, | |
"arc_easy": 0, | |
"arc_challenge": 0, | |
"sciq": 0, | |
"piqa": 0 | |
} | |
} |