minhleduc's picture
E2 run e2_ce_only_seed42 (epoch_4.pt)
69908be verified
Raw History Blame Contribute Delete
2.63 kB
{
"raw": {
"ece": 0.1403095155656338,
"brier": 0.050999487702119047,
"nll": 0.860319820090197,
"accuracy": 0.786,
"soft_accuracy": 0.5234679394811392,
"mean_confidence": 0.6456904814988375,
"score_mae": 0.21561100330349292,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.14761911233266192,
"brier": 0.06340317008623818,
"nll": 0.9857141292778154,
"accuracy": 0.7633333333333333,
"soft_accuracy": 0.4782372895628214,
"mean_confidence": 0.6157142215967178,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.1623956760764122,
"brier": 0.05286887642540137,
"nll": 1.0355604347772895,
"accuracy": 0.74625,
"soft_accuracy": 0.44410819893702863,
"mean_confidence": 0.5845614827796817,
"score_mae": 0.21561100330349292,
"n": 800
},
"noul": {
"ece": 0.10449459860722224,
"brier": 0.03610328702029015,
"nll": 0.5012713579864552,
"accuracy": 0.8616666666666667,
"soft_accuracy": 0.6745115767916043,
"mean_confidence": 0.7571720730264981,
"score_mae": null,
"n": 600
}
}
},
"post_temperature": {
"ece": 0.15913432231545446,
"brier": 0.04916353706188913,
"nll": 0.8560048570781946,
"accuracy": 0.786,
"soft_accuracy": 0.5143453535288572,
"mean_confidence": 0.6268656734824181,
"score_mae": 0.21544989785994403,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.16967813124259315,
"brier": 0.06114765697592588,
"nll": 0.9807129614241421,
"accuracy": 0.7633333333333333,
"soft_accuracy": 0.4674826178451379,
"mean_confidence": 0.5936552043259143,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.1808890850841999,
"brier": 0.05069461653213694,
"nll": 1.030125109362416,
"accuracy": 0.74625,
"soft_accuracy": 0.43563373532146216,
"mean_confidence": 0.565802447013557,
"score_mae": 0.21544989785994403,
"n": 800
},
"noul": {
"ece": 0.12017289320627847,
"brier": 0.03513797785418864,
"nll": 0.4991364163532853,
"accuracy": 0.8616666666666667,
"soft_accuracy": 0.6661569134891033,
"mean_confidence": 0.7414937779307366,
"score_mae": null,
"n": 600
}
}
},
"fitted_temperature": {
"type0_bucket1": 1.0989445447921753,
"type2_bucket0": 1.1291471719741821,
"type1_bucket1": 1.0874115228652954
}
}