minhleduc's picture
e2_ce_only_seed44: test_eval_v2.json
3631704 verified
Raw History Blame Contribute Delete
2.97 kB
{
"raw": {
"ece": 0.13831696626544,
"brier": 0.051284364823037706,
"nll": 0.8615161191206425,
"accuracy": 0.783,
"soft_accuracy": 0.5222538067623973,
"mean_confidence": 0.6452848957926035,
"mean_target_max": 0.6589462240487337,
"score_mae": 0.21527907053823583,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.15338748335838318,
"brier": 0.06259250025831231,
"nll": 0.986831498829027,
"accuracy": 0.765,
"soft_accuracy": 0.47661630041897296,
"mean_confidence": 0.6136187381048998,
"mean_target_max": 0.6345381752649943,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.1602796625718474,
"brier": 0.05328848799827483,
"nll": 1.0365782652422786,
"accuracy": 0.74375,
"soft_accuracy": 0.44316409012302754,
"mean_confidence": 0.5840223825722933,
"mean_target_max": 0.5928848401084542,
"score_mae": 0.21527907053823583,
"n": 800
},
"noul": {
"ece": 0.09865541060765583,
"brier": 0.0373040651541136,
"nll": 0.50278454458341,
"accuracy": 0.8533333333333334,
"soft_accuracy": 0.6733442686249812,
"mean_confidence": 0.7586344044407208,
"mean_target_max": 0.7714361180861791,
"score_mae": null,
"n": 600
}
}
},
"post_temperature": {
"ece": 0.1613641293048859,
"brier": 0.04948239230941744,
"nll": 0.8574540826445446,
"accuracy": 0.783,
"soft_accuracy": 0.5113731906339526,
"mean_confidence": 0.6228788203001022,
"mean_target_max": 0.6589462240487337,
"score_mae": 0.21538325707893818,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.19152938291430474,
"brier": 0.06042450366864311,
"nll": 0.9820380164931218,
"accuracy": 0.765,
"soft_accuracy": 0.45812685097257294,
"mean_confidence": 0.5765153206884861,
"mean_target_max": 0.6345381752649943,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.17983196027576925,
"brier": 0.05104208130190152,
"nll": 1.031392307500355,
"accuracy": 0.74375,
"soft_accuracy": 0.43448590984568,
"mean_confidence": 0.564741894416511,
"mean_target_max": 0.5928848401084542,
"score_mae": 0.21538325707893818,
"n": 800
},
"noul": {
"ece": 0.11253386884927749,
"brier": 0.03646069562687969,
"nll": 0.5009525156548867,
"accuracy": 0.8533333333333334,
"soft_accuracy": 0.6671359046796957,
"mean_confidence": 0.7467582210898399,
"mean_target_max": 0.7714361180861791,
"score_mae": null,
"n": 600
}
}
},
"fitted_temperature": {
"type1_bucket1": 1.0892750024795532,
"type0_bucket1": 1.1707441806793213,
"type2_bucket0": 1.0939185619354248
}
}