minhleduc's picture
e2_ce_only_seed42: test_eval_v2.json
0295eaa verified
Raw History Blame Contribute Delete
2.97 kB
{
"raw": {
"ece": 0.1398480086773634,
"brier": 0.05103822000877925,
"nll": 0.8604203677009791,
"accuracy": 0.7855,
"soft_accuracy": 0.5235573233515024,
"mean_confidence": 0.6458810180872678,
"mean_target_max": 0.6589462240487337,
"score_mae": 0.21578076076169964,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.14743219847480457,
"brier": 0.06339079040756132,
"nll": 0.9857434277236462,
"accuracy": 0.7633333333333333,
"soft_accuracy": 0.47836515128612517,
"mean_confidence": 0.6159011376897494,
"mean_target_max": 0.6345381752649943,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.16289205316454172,
"brier": 0.05298609249110598,
"nll": 1.0357897561136633,
"accuracy": 0.74625,
"soft_accuracy": 0.4441511295735836,
"mean_confidence": 0.584728074260056,
"mean_target_max": 0.5928848401084542,
"score_mae": 0.21578076076169964,
"n": 800
},
"noul": {
"ece": 0.10260182797908782,
"brier": 0.036088486300228206,
"nll": 0.5012714564613998,
"accuracy": 0.86,
"soft_accuracy": 0.6746244204541048,
"mean_confidence": 0.7573981569210688,
"mean_target_max": 0.7714361180861791,
"score_mae": null,
"n": 600
}
}
},
"post_temperature": {
"ece": 0.1588473957479,
"brier": 0.04917404281355277,
"nll": 0.8560326781123877,
"accuracy": 0.7855,
"soft_accuracy": 0.5143366709724069,
"mean_confidence": 0.6268481774032116,
"mean_target_max": 0.6589462240487337,
"score_mae": 0.2155858551512938,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.1697476014494896,
"brier": 0.06110521121062883,
"nll": 0.9806674488882223,
"accuracy": 0.7633333333333333,
"soft_accuracy": 0.4674839984625578,
"mean_confidence": 0.5935857443511486,
"mean_target_max": 0.6345381752649943,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.1827498009428382,
"brier": 0.05077533899652735,
"nll": 1.0302575784828514,
"accuracy": 0.74625,
"soft_accuracy": 0.4355850401893258,
"mean_confidence": 0.5657574540749192,
"mean_target_max": 0.5928848401084542,
"score_mae": 0.2155858551512938,
"n": 800
},
"noul": {
"ece": 0.11843508313099545,
"brier": 0.035107812839177244,
"nll": 0.4990980401759346,
"accuracy": 0.86,
"soft_accuracy": 0.6661915178596973,
"mean_confidence": 0.741564908226331,
"mean_target_max": 0.7714361180861791,
"score_mae": null,
"n": 600
}
}
},
"fitted_temperature": {
"type0_bucket1": 1.1001685857772827,
"type2_bucket0": 1.130467176437378,
"type1_bucket1": 1.0884613990783691
}
}