minhleduc's picture
e2_ce_only_seed43: test_eval_v2.json
09673cc verified
Raw History Blame Contribute Delete
2.97 kB
{
"raw": {
"ece": 0.1394936109185219,
"brier": 0.05230583777528644,
"nll": 0.8615829679872841,
"accuracy": 0.778,
"soft_accuracy": 0.5191233890913427,
"mean_confidence": 0.6385063585042954,
"mean_target_max": 0.6589462240487337,
"score_mae": 0.22387071973469574,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.15285709172487258,
"brier": 0.06384976085329451,
"nll": 0.9866861464145283,
"accuracy": 0.765,
"soft_accuracy": 0.47518984275559584,
"mean_confidence": 0.6121429173648357,
"mean_target_max": 0.6345381752649943,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.14815345995128154,
"brier": 0.05754133870081887,
"nll": 1.0408480976661667,
"accuracy": 0.72375,
"soft_accuracy": 0.4375544776860625,
"mean_confidence": 0.5755965473875403,
"mean_target_max": 0.5928848401084542,
"score_mae": 0.22387071973469574,
"n": 800
},
"noul": {
"ece": 0.11458379139502846,
"brier": 0.03378124679656848,
"nll": 0.49745961665486294,
"accuracy": 0.8633333333333333,
"soft_accuracy": 0.6718154839674632,
"mean_confidence": 0.7487495477994283,
"mean_target_max": 0.7714361180861791,
"score_mae": null,
"n": 600
}
}
},
"post_temperature": {
"ece": 0.1543409335911274,
"brier": 0.05096334562510441,
"nll": 0.8585448775477708,
"accuracy": 0.778,
"soft_accuracy": 0.5120342172160744,
"mean_confidence": 0.6236590646505356,
"mean_target_max": 0.6589462240487337,
"score_mae": 0.2241619166044984,
"n": 2000,
"by_type": {
"choice": {
"ece": 0.1725053222974141,
"brier": 0.062087079074434164,
"nll": 0.982777592378358,
"accuracy": 0.765,
"soft_accuracy": 0.46569606085618337,
"mean_confidence": 0.5924946775535742,
"mean_target_max": 0.6345381752649943,
"score_mae": null,
"n": 600
},
"score": {
"ece": 0.16392989568412306,
"brier": 0.05568060457957927,
"nll": 1.0367412413284183,
"accuracy": 0.72375,
"soft_accuracy": 0.4305423962138593,
"mean_confidence": 0.5598201040551066,
"mean_target_max": 0.5928848401084542,
"score_mae": 0.2241619166044984,
"n": 800
},
"noul": {
"ece": 0.12339126070340473,
"brier": 0.03354993356980818,
"nll": 0.49671701100965343,
"accuracy": 0.8633333333333333,
"soft_accuracy": 0.6670281349122524,
"mean_confidence": 0.7399420658747355,
"mean_target_max": 0.7714361180861791,
"score_mae": null,
"n": 600
}
}
},
"fitted_temperature": {
"type0_bucket1": 1.0873219966888428,
"type2_bucket0": 1.0724766254425049,
"type1_bucket1": 1.0740853548049927
}
}