minhleduc's picture
E2 run e2_laya_rlce_seed44 (epoch_4.pt)
0007194 verified
Raw History Blame Contribute Delete
2.63 kB
{
"raw": {
"ece": 0.13394594296813012,
"brier": 0.05498596685757368,
"nll": 0.8672405917616561,
"accuracy": 0.7715,
"soft_accuracy": 0.516161304473877,
"mean_confidence": 0.6375540485084057,
"score_mae": 0.22897809116868303,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.12707374493281046,
"brier": 0.06884223616547691,
"nll": 0.9961476571982105,
"accuracy": 0.7433333333333333,
"soft_accuracy": 0.4729881009707848,
"mean_confidence": 0.6162595846255621,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.17143568731844427,
"brier": 0.057610363487638094,
"nll": 1.0435749628953637,
"accuracy": 0.73875,
"soft_accuracy": 0.4323799050040543,
"mean_confidence": 0.5673143507540226,
"score_mae": 0.22897809116868303,
"n": 800
},
"noul": {
"ece": 0.09800974751512209,
"brier": 0.037630502042917915,
"nll": 0.5032210314801584,
"accuracy": 0.8433333333333334,
"soft_accuracy": 0.6710430406033993,
"mean_confidence": 0.7525014427304267,
"score_mae": null,
"n": 600
}
}
},
"post_temperature": {
"ece": 0.1504889203310013,
"brier": 0.0535475225614009,
"nll": 0.864496507877484,
"accuracy": 0.7715,
"soft_accuracy": 0.5082126140743494,
"mean_confidence": 0.6210110761225224,
"score_mae": 0.23021191493957305,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.15238359754284223,
"brier": 0.06618621635801294,
"nll": 0.9919972387328744,
"accuracy": 0.7433333333333333,
"soft_accuracy": 0.460888055190444,
"mean_confidence": 0.5909497653941314,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.18502540819346902,
"brier": 0.05639494055324576,
"nll": 1.040980128934607,
"accuracy": 0.73875,
"soft_accuracy": 0.4262336232326925,
"mean_confidence": 0.5537245621159673,
"score_mae": 0.23021191493957305,
"n": 800
},
"noul": {
"ece": 0.10953281487027804,
"brier": 0.037112271442329016,
"nll": 0.501684282279263,
"accuracy": 0.8433333333333334,
"soft_accuracy": 0.664842494080464,
"mean_confidence": 0.7407877388596534,
"score_mae": null,
"n": 600
}
}
},
"fitted_temperature": {
"type1_bucket1": 1.0628252029418945,
"type0_bucket1": 1.108896255493164,
"type2_bucket0": 1.0960332155227661
}
}